Executive Summary
Matterly is a self-hosted legal practice management SaaS with multi-tenant architecture, built on Laravel 12, React 19, TypeScript and Inertia.js. Two groups stand to benefit most: operators who want to run many independent law firms from one installation, and firms that prefer a single system for their website, intake, matters, forms, billing and client messaging. This article studies it as an AI legal practice management SaaS, concentrating on how it is put together.
Three traits set it apart technically. The first is tenancy: a single installation serves separate firms under a Super Admin layer, so the boundary between firms becomes the central engineering question. One installation does not mean one shared pool of data. The second is the Smart Form, a document that carries its legal text, inline fields and electronic signature together, and that lets staff send back individual fields for correction. The third is the matter, which serves as the anchor for tasks, forms, invoices, messages and the progress a client sees.
AI sits on top as an option instead of a foundation. The practice-management core runs without it, and AI features need a configured OpenAI or OpenAI-compatible provider, with staff review built into the workflow. Everything here rests on the verified feature baseline dated 2 October 2026.
At a Glance
| Item | Detail |
|---|---|
| Product | Matterly: AI Law Firm Management SaaS with Smart Forms |
| Category | Self-hosted, multi-tenant legal practice management SaaS |
| Stack | Laravel 12, React 19, TypeScript, Inertia.js, Tailwind CSS |
| Runtime | PHP 8.2+, with MySQL 8+ or MariaDB 10.6+ |
| Deployment | Self-hosted. Browser-based installer. A single cron job |
| Primary users | Platform owner (Super Admin), Firm Admins, staff and paralegals, clients, public visitors |
| Multi-tenancy | One installation, several independent firms, plus impersonation and suspension controls |
| AI | Optional. Works through an OpenAI or OpenAI-compatible API |
| Payments | Stripe, Razorpay, PayPal, LigdiCash and pay-at-office. SaaS plan billing recurs through Stripe |
| Communications | SMTP email, SMS (Twilio, MSG91, Africa's Talking), WhatsApp workflows, website chat, client inbox |
| Core workflow | Visitor → consultation → matter → Smart Form → signature → review → billing → client portal |
How to Read This Case Study
Claims in this article come in three kinds, each marked so you can separate verified facts from interpretation.
- Confirmed capability: a feature listed in the verified baseline, stated without a label.
- Architectural reading: an inference about what a capability implies for system design. It says nothing certain about the private source code.
- Implementation consideration: advice for teams planning a production rollout.
The baseline is silent on database schemas, how tenancy is enforced, queue systems, hosting providers and certifications, so this article stays silent too. Where such details matter, it names the concern and suggests checking the source code.
Why Fragmented Legal Software Fails
Many practices stitch together a website builder, a booking tool, an intake spreadsheet, a case-management application, emailed attachments, a standalone e-signature service, an invoicing package, a messaging tool and a portal added late. None of these tools is the problem. The trouble sits in the gaps between them.
- Re-keyed data: client details get typed again for booking, intake and invoicing.
- Split intake: a consultation request sits in one system while the case lives in another.
- Email-based paperwork: forms travel as attachments that are filled in, scanned, sent back and filed manually.
- Chasing by hand: staff request documents and send reminders because nothing connects a case stage to its next step.
- No single view: tasks, deadlines, forms, invoices and messages are spread across tools.
- Clients left guessing: people call the office to ask where their matter stands.
Serving many firms compounds all of this. An operator building a legal SaaS has to solve each problem once and then keep every firm’s version apart, with its own staff, services, fees, stages, fields, website and settings. Without a purpose-built platform, the operator ends up running separate installations or gluing together tools never meant to be shared between tenants. That is the territory of multi-tenant SaaS architecture.
The Unified Multi-Tenant Solution
Matterly answers fragmentation by treating the practice as one connected system. The verified feature set breaks down into fifteen domains that act on one another.
| Domain | Role |
|---|---|
| Platform administration | The Super Admin manages firms, plans, languages and platform-wide monitoring |
| Firm administration | Practice areas, services, fees, staff, stages and settings |
| Matter management | The operational core that other domains attach to |
| Client management | People and organisations linked to matters |
| Smart Forms | Document-style data collection with signature |
| Document and signature workflow | Signed PDFs filed against the matter |
| Calendar and tasks | Hearings, deadlines, meetings, consultations, tasks and due dates |
| Billing and payments | Invoices, payments, expenses and balances |
| Communications | Email, SMS, WhatsApp workflows, chat and client inbox |
| Client portal | A controlled, client-facing view |
| Firm website | Public site, practice-area pages and consultation booking |
| AI services | Optional generation, assistance, insights and agents |
| SaaS subscriptions | Plans, limits, credits and recurring billing |
| Reporting | Operational and financial views for firms, revenue analytics for the operator |
| Support | Operator-side help, including controlled impersonation |
The Matter-Anchored Workflow Model
The Matter-Anchored Workflow Model is a lens for reading the verified features. It is an architectural interpretation, not a description of Matterly’s internal database design. Under this lens the matter is the anchor: every other domain either supplies it with data, reads from it, or reacts when it changes.
Value comes from the links in this chain, not from any single module. A submitted Smart Form is filed on the matter. An invoice belongs to the matter. A stage change can spawn tasks and messages on that same matter. The portal then presents the client with a restricted view of it. Each link removes a manual handoff that a fragmented stack would leave in place.
High-Level Architecture of an AI Legal Practice Management SaaS
At a conceptual level, the feature set points to a layered structure.
The platform owner works above the tenant boundary. Each firm workspace contains the practice domains, and outside services connect to it as integrations.
| Integration group | Named in the baseline | Role |
|---|---|---|
| AI | OpenAI or an OpenAI-compatible API | Generation, assistance, insights, chat, agents |
| Payments | Stripe, Razorpay, PayPal, LigdiCash | Consultation and invoice payments, platform subscriptions |
| SMS | Twilio, MSG91, Africa's Talking | Two-way SMS and reminders |
| SMTP | Notifications, form delivery, reminders | |
| WhatsApp workflows | Messaging where the integration is set up |
Architectural reading: the integration list indicates that outside services are swappable edges around a stable core. Each integration relies on configuration by the operator or the firm. None is active automatically, and the core practice features do not hinge on a single vendor.
Multi-Tenant Architecture: One Installation, Many Independent Firms
Matterly is multi-tenant: a single installation hosts several independent firms beneath a Super Admin layer. Every firm runs as its own legal practice, with its own:
- configuration, such as practice areas, services, fees, stages and custom fields
- users and permissions
- clients and matters
- website, branding and domain or subdomain
- billing and payment settings
- AI usage and credits
One installation is not one shared data pool. An installation describes how the software is deployed. Isolation describes how data is kept apart. The baseline confirms that firms are independent but does not say how that separation is enforced, so this article offers no claim about the mechanism. Readers who want general background on the design space can explore multi-tenant database architecture.
Four concerns follow from this model.
- Tenant context. Each request, public booking and scheduled job has to resolve to the right firm before it reads or writes anything.
- Data isolation. One firm’s records must never surface in another firm’s workspace. For legal software this is the most important property, since the data is confidential by nature.
- Configuration isolation. A family-law practice and a corporate practice on the same installation should not be forced to share a pipeline, a field list or a payment setting.
- Platform-level control. The Super Admin sits above tenants, which is what makes plan management, language management and operator-level monitoring possible.
Two Super Admin capabilities need their own mention. Impersonation lets the operator view a firm’s workspace the way that firm sees it, replacing guesswork during support. Suspension lets the operator switch off a single firm, for example over unpaid fees, while unrelated tenants carry on. Both only work when firms are cleanly separable units.
Validate before production. Test tenant isolation against the real implementation before hosting any firm’s data. Treat impersonation as a privileged action governed by a written policy.
Roles and Permissions
Legal software needs more than an Admin and a User. People carry different responsibilities, and many should see only a slice of a matter.
| Role | Responsibility | Access boundary |
|---|---|---|
| Platform Owner / Super Admin | Runs the SaaS: plans, firms, languages, credits, revenue, support | Platform level, above every firm |
| Firm Admin | Sets up and runs a single firm | That firm's workspace, settings and billing |
| Staff / Paralegal | Handles matters, tasks, forms, calendar and communications | Narrowed by granular permissions within the firm |
| Client | Follows their own matters, invoices and balance | Their own portal information only |
| Public Visitor | Browses the website, chats and books a consultation | Public pages and the booking flow |
The baseline includes granular permissions and permissions at the level of matter-workspace tabs. Architectural reading: this indicates authorisation that controls what someone can see inside a matter, beyond which modules they may open.
Three everyday cases show why that matters.
- A receptionist may need the calendar but not billing.
- A paralegal may need forms and tasks but not financial information.
- A client should see only their own portal information.
Tab-level control lets a firm draw those lines without inventing a new role for every combination. The client-facing side is covered in the client portal section below.
Matter-Centred Workflow
The matter workspace is the operational center of the platform. It ties together the client, tasks, deadlines, forms, billing, communications, activity, custom fields, practice area and stage. Billing, tasks, forms and communication are not treated as separate applications. They share one matter context.
From there, a staff member can go from case status to open work, submitted forms, billing details and client messages without changing systems. Architectural reading: building around the matter sidesteps the synchronisation headaches of fragmented tools, because there is one record to update and one record to read. For the wider category, see legal case management software.
Kanban Boards and Workflow Automation
Matters travel through stages on a Kanban board. Desktop users drag matters from column to column. On mobile, a “Move to” action takes the place of dragging, which keeps the board practical on a phone. Stages can be color-coded, and pipelines can be defined per practice area, so a property transaction and a litigation matter need not follow one shared path.
Every stage has two names.
| Name | Audience | Example |
|---|---|---|
| Internal stage name | Firm staff | "Awaiting counsel review" |
| Client-facing stage name | Client, in the portal | "We are reviewing your documents" |
Staff keep the vocabulary they need, while clients see progress in plain language without internal detail.
Moving a matter into a stage can create tasks, deliver Smart Forms, send client messages and set off other follow-up. That changes the board from a visual tracker into an operational workflow engine, one that causes work instead of merely displaying it. The example below is an illustrative workflow built on those automation capabilities.
Each matter that reaches that stage gets identical preparation, no matter which staff member moved it. Follow-up that follows rules becomes configuration rather than something people must remember. For related patterns, see workflow automation.
Custom Matter Data
Practice areas gather different facts. Family, immigration, corporate, property and litigation matters share no single data shape. A fixed schema either leaves out fields some firms need or crowds every matter with fields most firms never use.
Matterly offers custom case fields, with several field types, filtering, pinned fields and configuration specific to each practice area. Pinned fields keep the key facts in view inside the matter workspace, and filters help staff locate matters by those values.
The important link is Smart Form to matter field mapping. Approved information from a submitted form can fill matter custom fields, which spares staff from retyping a client’s answer into a case record.
AI Smart Forms: Document-Centric Data Collection
A Smart Form is a document that holds legal text, structured fields and an electronic signature together. After multi-tenancy, it is the platform’s most distinctive technical feature.
An ordinary web form pulls the questions away from their meaning. The client answers a list, and the firm assembles a document later. A Smart Form keeps the two together.
The client reads the wording and fills each field where it appears, next to the clause it relates to, then signs. Architectural reading: collecting data inside the document can make the experience clearer for the client, since each answer arrives with its context, and it can shorten the road to a filed record because the result is already the document the firm wants.
Staff can produce a Smart Form with AI or build one by hand. The AI route is assistive, with a person involved at each step.
A staff member describes the document in everyday language. AI proposes wording and fields. Staff review and edit the draft in the visual editor, and human approval stays part of the workflow before anything reaches a client. Generated text is never presented as automatically correct, legally valid or ready to use, and a Smart Form does not itself give legal advice.
AI takes part only at the start of this pipeline. Completing fields, signing, producing the PDF and filing it are ordinary application behavior. For related reading, see AI Smart Forms and AI document automation.
Smart Form Builder
The builder is a visual, drag-and-drop editor where legal text and input fields share one document. It runs on mobile as well as desktop, and any field can be set as required. Manual and AI-assisted creation use the same editor, so AI output arrives somewhere staff can read and change it. The builder offers nine field types.
| Field type | Typical use |
|---|---|
| Short text | Names, reference numbers |
| Long text | Statements, narrative answers |
| Number | Amounts, quantities |
| Date | Birth dates, effective dates |
| Dropdown | Picking from a fixed list |
| Single choice | Choosing exactly one option |
| Checkbox | Confirmations, acknowledgements |
| File upload | Supporting documents |
| Signature | Capturing an electronic signature |
Secure Sharing and Client Completion
Each form goes out through a randomly generated 32-character secure link. Access can be guarded with an optional PIN, limited by an expiry date, and switched on or off whenever needed. Forms can be sent by email, one matter can hold several forms, and every submission is filed automatically under its matter. That replaces the loop of attaching, downloading, scanning and filing documents manually.
The client’s path is brief.
- The client receives the secure link and opens the document.
- They read the legal text and fill in fields inline, with required fields highlighted.
- They sign electronically and submit.
- They receive a signed PDF receipt.
The link and, where set up, the PIN control access. The form also works on mobile, which matters because many clients will open it on a phone.
Implementation consideration: link-based sharing, a PIN, expiry and the option to disable access are useful safeguards, though they are safeguards and not guarantees. A firm should decide when a PIN is required, how long links stay valid and how links are delivered to clients.
Field-Level Corrections
Review is where many form workflows stall. A client submits a document, staff spot one wrong answer, and the usual fix is to return the entire form. Matterly takes a narrower route with field-level corrections.
Clients see only the fields that need attention. Staff get a changes view comparing before and after values, so they check the difference instead of rereading everything. Architectural reading: this implies submissions are kept in a form that tracks individual field values over time, instead of as an opaque finished file. Once accepted, approved values can flow into matter custom fields through the mapping described earlier.
AI as an Optional Service Layer
AI in Matterly is an optional service layer placed over a core application that works without it.
| Layer | Contents |
|---|---|
| Core application | Matters, clients, tasks, calendar, billing, manual Smart Forms, permissions, portal, website, communications |
| AI service | Smart Form generation, website generation, section and SEO writing, translation, practice descriptions, theme suggestions, email writing, review responses and insights, marketing, fee advice, revenue insights, Ask Your Data, website chatbot, AI Staff |
A firm can handle matters, billing, calendar, portal, website and manual Smart Forms without ever connecting an AI provider. AI features do need provider access. Keeping AI optional lets an operator launch without taking on AI costs, and a provider outage, or a choice to turn AI off, does not stop the practice.
One assistive principle runs through the whole layer: AI proposes, and staff review and edit. It applies to website content, SEO writing, translation, review replies, client updates, insights, fee advice, revenue insights and Ask Your Data. The baseline presents none of these outputs as legal advice. For background on the pattern, see human-in-the-loop AI.
AI Provider Architecture
The baseline supports OpenAI and any OpenAI-compatible API. It names no other providers, and this article does not claim that arbitrary models will work.
Architectural reading: support for an OpenAI-compatible interface suggests a provider abstraction, in which application features call one consistent AI interface instead of a single vendor. In concept, that brings provider flexibility, less vendor lock-in, one place for configuration and simpler provider changes. How well any particular compatible service handles a given feature should be tested first.
AI Credits
AI credits give the operator a way to meter and cap AI spend for each firm.
Each firm holds a balance of AI credits. Requests draw it down, usage is recorded, and the platform owner manages and monitors credits across the platform. AI Staff agents additionally run under a daily budget. The baseline does not say how credits are calculated, so the article makes no claim about token pricing or credit formulas.
In a multi-tenant product, the operator pays for AI usage unless it is metered, so these controls are a commercial necessity.
- Cost control: one heavy tenant cannot quietly absorb a disproportionate share.
- Tenant-level metering: usage can be traced to a specific firm.
- Pricing tiers: AI usage can be bundled into a plan or sold as a top-up.
- Protection against excessive consumption: limits give firms and the operator predictable boundaries.
Credits sit beside SMS credits and plan limits in the commercial control layer, discussed under SaaS subscription architecture below.
AI Staff
AI Staff are agents that handle recurring administrative work. The baseline confirms four categories.
- Review replies: drafts responses to client reviews.
- Client updates: drafts status messages for clients.
- Daily task planning: proposes a plan for the day’s work.
- Revenue notes: summarises financial observations.
Their defining feature is the draft-first model.
Agents write drafts. Drafts wait in an approval inbox. A person reads, then sends them, so nothing reaches a client simply because an agent produced it. This is important for legal-business communication: a review reply is public, a client update can shape expectations, and even routine wording carries professional weight. These agents assist with administrative and business tasks. They are not autonomous, they make no legal decisions, and they are not presented as giving legal advice. For related background, see AI agents.
AI Website Chat
A public chatbot can answer visitors’ questions from the firm’s knowledge base and help with consultation booking.
The chatbot draws on content the firm controls, so the firm decides what it knows. The baseline gives no detail on a retrieval or indexing technique, and this article assumes none. Implementation consideration: read through the knowledge base carefully and try the chatbot on sensitive questions before putting it on a live site.
Website and CMS
Each firm can publish its own website, with branding, pages, menus, themes, practice-area pages, SEO settings, a custom domain or subdomain and multilingual content. The baseline lists six templates and a block editor.
The architectural gain is that the website and the practice-management data sit in the same firm context. A standalone website builder cannot see a firm’s services, fees or practice areas. Inside the platform, a visitor reading a practice-area page can book a consultation that creates records in that firm’s own workspace. The baseline says nothing about how the CMS is built, so the article adds nothing on that point. Operators who plan to resell the experience under their own brand can read more on SaaS white labeling.
Consultation Booking
The booking flow converts a visitor into a client record.
The baseline confirms several supporting details: a payment hold while checking out, a message to the visitor when payment fails, SMS consent capture and matter lookup for returning clients. A firm can take payment online or let visitors choose to pay at the office.
Architectural reading: the payment hold and the failure message suggest the flow has defined behavior for the stretch between “requested” and “paid”. The baseline does not describe the authorization or capture mechanism underneath, and it names no calendar provider, so the article does neither.
Billing and Payments
Billing travels with the matter instead of sitting beside it.
The baseline covers invoices, payments, expenses, outstanding balances, payment reminders and overdue reminders. Because invoices belong to the matter, staff see fees, payments and balances alongside tasks and forms, and clients see their own invoices and balance in the portal.
The supported gateways are Stripe, Razorpay, PayPal and LigdiCash, plus a pay-at-office option for offline payment. Webhook-based payment workflows are included, which is the standard way for a payment provider to report a result back to an application. Stripe’s webhook guide and the Razorpay documentation cover the provider side of those mechanics. For the wider topic, see payment gateway integration.
Implementation consideration: gateway availability varies with configuration and with regional and provider coverage. Confirm that the chosen gateway operates in the firm’s country before launch.
SaaS Subscription Architecture
Matterly bundles the commercial control layer a platform operator relies on.
The baseline lists subscription plans, recurring Stripe billing, saved cards, platform invoices, coupons, multiple currencies, feature flags, plan limits, AI credits, SMS credits and revenue analytics. It sets out no subscription tiers or prices, and none are implied here.
Feature flags and limits sit at the heart of any SaaS. They let the operator build tiers by switching features on or off, capping usage and metering AI and SMS. Without them, every firm would get the same capabilities at the same price, and heavy usage would face no ceiling. Revenue analytics show the operator how the business is performing. For broader coverage, see SaaS billing.
Communications and Reminders
Client communication runs through SMTP email, SMS, WhatsApp workflows, website chat and a client inbox. SMS can go through Twilio, MSG91 or Africa’s Talking.
The SMS layer provides two-way messaging, delivery status, STOP and START handling, quiet hours and message logs. These controls matter because messaging clients carries expectations around consent and timing: STOP and START let people manage their own subscription, and quiet hours keep texts from arriving at awkward times.
Provider credentials can be managed in two ways.
| Model | How it works | Who holds the provider account |
|---|---|---|
| Platform provider | The operator keeps the provider account and resells credits | Operator |
| Bring-your-own provider | The firm supplies its own provider credentials | Firm |
Architectural reading: supporting both models suggests messaging is decoupled from the provider, so the same firm-facing features work whichever credential source is active. SMS depends on provider configuration. WhatsApp workflows depend on the supported integration setup and should not be assumed to run without it. For background, see SMS API integration, WhatsApp Business integration and the Twilio documentation.
Automated reminders cover hearings, consultations, invoice due dates and overdue payments, and can go out by SMS, WhatsApp or email. These tasks are repetitive and time-sensitive, and need no judgement, so they suit automation. Once configured, they cut down the forgotten follow-up behind missed hearings and unpaid invoices.
Client Portal
The client portal is a controlled visibility layer, not a duplicate of the staff dashboard. Clients activate an account, then see their matters, progress, invoices, balance and relevant information, and they communicate through the supported channels.
Client-facing stage names are where the portal pays off. A client reads “We are reviewing your documents” instead of an internal pipeline label, which answers the most common question, where does my matter stand, without a phone call. Architectural reading: because the portal is a deliberately narrow projection of the matter, its permission design is as important as its interface. For the engineering side, see client portal development.
Calendar
A single practice calendar brings together hearings, deadlines, meetings, consultations, tasks and invoice due dates. It offers agenda, week, month and team views, allows drag-and-drop rescheduling and can be filtered by practice area.
A unified calendar improves visibility. When hearings, deadlines and consultations live in separate calendars, clashes are easy to overlook. One calendar with team and practice-area views can help reduce that risk and shows workload and commitments side by side.
Multilingual and RTL Architecture
The interface is available in English, French, Spanish and German, with a partial Hindi starter translation. The Super Admin can add more languages, and AI translation is available. Firm website pages can be multilingual.
Smart Forms handle right-to-left text, including Arabic, Hebrew and Urdu. RTL matters for document-oriented forms because direction touches more than alignment. Text flow, field placement, punctuation and signature areas all depend on it, and a form that reads awkwardly in the client’s language defeats the purpose of a document-centric experience.
Technical SEO for Multi-Tenant Legal Websites
The baseline lists multilingual indexing, hreflang, robots.txt, sitemap.xml, llms.txt, structured data, legal-service schema, opening hours, practice areas, FAQs, documentation and social previews.
Multi-tenant SEO brings a particular difficulty. Many firms share one installation and each may use its own domain or subdomain, so each site must present its own crawl rules, sitemap, language alternates and structured data instead of inheriting the platform’s. If those signals bleed across tenants, indexing gets muddled and firms lose their own identity in search.
Structured data for services, opening hours and practice areas, kept separate for each firm, helps search engines and AI answer systems grasp what each firm does. Structured data does not guarantee rich results, and none of these features guarantees rankings.
Two references are worth bookmarking. Google documents localized versions and hreflang and sitemaps. The llms.txt proposal outlines an emerging convention for helping AI agents read a website. It is a proposal, not a Google requirement.
Implementation consideration: test each firm’s canonical, language-alternate and sitemap behavior on a live domain before launch. For the wider topic, see technical SEO for SaaS.
Security Considerations for Legal SaaS
The following are architectural considerations based on the verified feature baseline. They are not certifications or compliance claims.
From an architectural perspective, the feature set shows attention to these concerns:
- Tenant isolation between firms, discussed earlier and the property most worth verifying.
- Granular permissions and matter tab permissions, which narrow what staff can see inside a matter.
- Secure Smart Form links, with an optional PIN, expiry and a switch to disable access.
- Controlled client portal access, limited to the client’s own information.
- Payment-provider workflows, handled through the configured gateways.
- Communication consent, through SMS consent capture, STOP and START, and quiet hours.
- Activity visibility on matters, so staff can see what has happened.
The verified baseline does not establish an encryption implementation, penetration testing, a security certification, a compliance certification or regulatory compliance. None of these should be assumed.
A production deployment should plan for:
- backups and recovery
- access control and credential management
- hosting location
- an update policy
- monitoring
- incident response
- data retention
- a privileged-access policy for impersonation
Legal organisations should weigh their own regulatory, privacy and security requirements before deployment and check any security claim against the source code.
Deployment
Matterly is self-hosted. The advantage is operator-controlled deployment: one installation, direct ownership of the infrastructure and an environment you can configure. The burden is just as direct. The operator is responsible for infrastructure, updates, backups, security hardening, monitoring and everyday maintenance. Self-hosting is not automatically safer than a managed service. It transfers responsibility to the operator. For the model in general, see self-hosted SaaS.
A browser-based installer checks the server’s requirements, writes the configuration, creates the database tables, optionally loads demo data and creates the Super Admin account. A deployment needs a public document root, writable directories and one cron job to run scheduled functions.
| Component | Version or note |
|---|---|
| PHP | 8.2+ |
| Laravel | 12 |
| React | 19 |
| TypeScript, Inertia.js, Tailwind CSS | Used |
| Database | MySQL 8+ or MariaDB 10.6+ |
Architectural reading: a single cron job suggests scheduled work such as reminders is driven from one scheduler entry, which keeps the operating requirement small. The baseline lists no other infrastructure components, and this article adds none. Teams familiar with the stack can find background in Laravel SaaS development and React SaaS development. Official references: the Laravel documentation, the React documentation, Inertia.js and the Tailwind CSS documentation.
End-to-End Workflow
The sequence below is an illustrative workflow assembled from confirmed capabilities. It does not describe internal processing.
| Phase | What happens | Architecture behind it |
|---|---|---|
| 1. Platform to firm | The operator creates a plan and onboards a firm. The firm configures practice areas, services, fees, staff, permissions, website, payments and messaging | Plans, feature flags and limits set what a firm may use. Firm setup is configuration isolation in practice |
| 2. Visitor to booking | A visitor requests a consultation and pays online or opts to pay at the office | Website and booking share the firm's services and fees. The payment hold and offline option cover the gap between request and payment |
| 3. Booking to matter | A matter is created and enters its practice-area pipeline. Stage automation creates tasks, messages or forms | The matter becomes the anchor, and stages turn a status change into work |
| 4. Document to signature | Staff generate or build a Smart Form. The client opens the secure link, completes fields and signs. A signed PDF is produced | One document structure serves AI-assisted and manual creation, and link controls govern access |
| 5. Review to data | Staff are notified, review the submission and either accept it or request field-level corrections. Approved values can fill custom fields | A controlled loop turns a document into structured matter data |
| 6. Ongoing operation | Tasks, deadlines, hearings, invoices, payments and communications continue. The client follows progress in the portal. Reminders run where configured | Calendar, billing, reminders and portal all read from the same matter context |
| 7. Oversight | The firm reviews operational and financial reports. The platform owner monitors subscriptions, credits, revenue and support | Firm reporting and platform analytics sit on separate layers |
Feature-to-Business-Value Table
Each table pairs a capability with the operational problem behind it and the business effect it can support. These are potential outcomes, not promised results.
For the Platform Operator
| Technical capability | Operational problem | Business impact |
|---|---|---|
| Multi-tenancy | One deployment needed per firm | Lets a single installation serve many firms |
| Super Admin impersonation | Firm-specific issues are hard to reproduce | Supports quicker, more accurate support |
| Firm suspension | One firm's status can disturb the platform | Supports contained operator control |
| Feature flags and plan limits | No way to tier or meter usage | Creates a foundation for tiered plans |
| Recurring Stripe billing, coupons, currencies | Subscriptions collected by hand | Supports recurring revenue in several currencies |
| AI and SMS credits | Open-ended AI and messaging costs | Supports cost management per firm |
| Revenue analytics | Little visibility into the SaaS business | Improves operator decisions |
| Browser installer, one cron job | Complicated setup | Can lower deployment effort |
For Firm Operations
| Technical capability | Operational problem | Business impact |
|---|---|---|
| Practice-area pipelines | Cases progress unevenly | Supports standardised workflows |
| Stage automation | Repetitive manual follow-up | Reduces manual work |
| Custom case fields | Rigid case schemas | Enables flexibility by practice area |
| Smart Form to matter field mapping | Answers retyped into case records | Reduces duplicate data entry |
| Granular and tab-level permissions | Overly broad access inside a matter | Supports safer delegation |
| Unified calendar | Hearings and deadlines scattered | Improves visibility of commitments |
| Automated reminders | Reminders sent manually | Reduces administrative effort |
| Integrated billing | Invoices detached from matters | Gives a clearer financial picture per matter |
| Multi-gateway payments | Few payment options | Widens payment coverage, subject to configuration |
| AI Staff with approval inbox | Recurring drafting and preparation | Supports staff productivity while keeping review |
For the Client Experience
| Technical capability | Operational problem | Business impact |
|---|---|---|
| Document-style Smart Forms | Questionnaires cut off from legal wording | Improves client comprehension |
| Secure links, PIN, expiry | Documents passed around by email | Supports controlled form delivery |
| Electronic signature, signed PDF | Print, sign, scan, send | Shortens document completion |
| Field-level corrections | Entire form returned for one error | Shortens review loops |
| Client portal, client-facing stage names | Status enquiries by phone and email | Improves client visibility |
| Multilingual and RTL support | Language barriers in forms and sites | Supports wider client reach |
| AI website chat with booking help | Slow first contact | Can reduce friction at first contact |
AI vs Non-AI Capability Matrix
Core matter management, billing, calendar and manual Smart Forms work without AI. AI generation needs a configured OpenAI or OpenAI-compatible provider. Payments rely on configured gateways, SMS on messaging providers and WhatsApp on the supported integration setup.
No AI Required
| Capability | External provider? | Purpose |
|---|---|---|
| Matters, clients, Kanban, stages, custom fields | No | Central operational record |
| Tasks and unified calendar | No | Scheduling and deadlines |
| Stage automation | Depends on the channel it triggers | Triggered tasks, forms, messages |
| Manual Smart Form builder | No | Legal text with input fields |
| Electronic signature and signed PDF | No | Capture signatures, produce documents |
| Field-level corrections | No | Targeted review and resubmission |
| Billing, invoices, expenses | No | Fees and balances |
| Online payments | Yes: Stripe, Razorpay, PayPal or LigdiCash, as configured | Collect payments |
| Pay-at-office | No | Offline payment option |
| Client portal | No | Client visibility |
| Public website and block editor | No | Firm web presence |
| Consultation booking | Payment gateway, if online payment is used | Visitor to client |
| Roles and permissions | No | Access control |
| SaaS plans, feature flags, coupons | Stripe for recurring billing | Operator business model |
| Yes: SMTP | Notifications and form delivery | |
| SMS and two-way SMS | Yes: Twilio, MSG91 or Africa's Talking | Client messaging |
| WhatsApp workflows | Depends on the supported integration setup | Client messaging |
| Reminders | Depends on the channel provider | Hearing, consultation and invoice reminders |
| Reporting and revenue analytics | No | Operational and financial views |
AI Required
Each item below needs a configured AI provider, and AI credits meter its usage.
| Capability | Purpose |
|---|---|
| AI Smart Form generation | Draft wording and fields from a prompt |
| AI website and content generation | Draft pages and sections |
| AI SEO writing | Draft SEO copy |
| AI translation | Translate interface and website content |
| AI website chatbot | Answer visitor questions from the firm's knowledge base |
| AI Staff | Draft review replies, client updates, task plans and revenue notes |
| AI insights (review and revenue) | Analyse reviews and revenue |
| Ask Your Data | Query firm data in plain language |
| AI fee advice | Offer fee-related suggestions |
Personas and Ideal Use Cases
Matterly is a potential fit for certain operators and practices, not for every firm.
| Persona | Primary needs | Relevant features |
|---|---|---|
| SaaS Operator | Launch and monetise a legal product for many firms | Multi-tenancy, plans, feature flags, credits, revenue analytics, impersonation, suspension |
| Law Firm Owner | Consistent workflows, financial visibility, controlled costs | Pipelines, billing, reports, AI controls |
| Office Manager | Smooth daily operations | Calendar, tasks, reminders, communications, permissions |
| Partner | Oversight of matters and revenue | Matter workspace, reports, revenue insights, team calendar |
| Paralegal | Efficient preparation and document handling | Smart Forms, custom fields, tasks, stage automation |
| Client | Clear progress and simple paperwork | Mobile Smart Forms, portal, invoices, reminders |
| Public Visitor | Finding a firm and booking quickly | Multilingual site, AI chat, booking, payment options |
The platform is particularly useful where an operator wants one installation for many firms, or where a practice wants digital intake and configurable workflows. Relevant settings include:
- multi-tenant legal SaaS operators and legal-tech agencies building white-label products
- independent law firms, advocate offices and legal consultancies
- multi-office practices that depend on team calendars and permissions
- organisations that require self-hosted deployment
- multilingual practices, including those serving right-to-left languages
Key Engineering Challenges
Each challenge follows one pattern: the problem, the design response visible in the feature set, and why it matters.
1. Tenant isolation. Problem: several firms must work independently on one installation. Design response: firm-scoped administration, users, clients, matters, billing, website and AI usage beneath a Super Admin layer. Why it matters: confidential legal data cannot cross firm lines, and the enforcement mechanism needs independent verification.
2. Document-oriented forms. Problem: structured input has to sit comfortably inside legal text. Design response: one document structure holding legal text and nine field types. Why it matters: clients answer in context, and the output is already a document.
3. AI reliability. Problem: generated content can be wrong. Design response: AI output lands in the same visual editor used for manual forms. Why it matters: staff can inspect and change every draft before a client sees it.
4. Human approval. Problem: automation should help staff without taking over. Design response: AI Staff write drafts that wait in an approval inbox. Why it matters: a person stays responsible for public and client-facing wording.
5. Workflow flexibility. Problem: practice areas follow different paths. Design response: practice-area pipelines, custom stages and custom fields. Why it matters: firms shape the system rather than bending to it.
6. Client experience. Problem: clients should complete legal documents without wrestling with software. Design response: secure links, inline completion, highlighted required fields, mobile support and a signed PDF receipt. Why it matters: the most important step in intake does not stall on usability.
7. Payment state management. Problem: online bookings can end up incomplete. Design response: a payment hold during checkout, failure feedback and a pay-at-office option. Why it matters: the booking flow has defined behavior between request and payment.
8. Communication provider abstraction. Problem: the platform must serve both centrally managed and firm-owned messaging accounts. Design response: platform-resold credits and firm-owned credentials behind one set of messaging features. Why it matters: the operator keeps a way to earn from messaging and firms keep their own contracts.
9. Usage-based AI economics. Problem: AI costs rise with usage. Design response: firm-level credits, usage tracking, platform monitoring and a daily AI Staff budget. Why it matters: spend stays attributable and bounded.
10. Multi-domain websites. Problem: many firms need separate web presences on one platform. Design response: per-firm sites with custom domains or subdomains and their own SEO settings. Why it matters: crawl, language and structured-data signals must not leak between tenants.
Architectural Design Principles
These principles are inferred from observable capabilities. They are not claims about private implementation details.
| Principle | What the feature set suggests |
|---|---|
| Tenant isolation | Firms are separate units with their own staff, clients, settings and usage |
| Modular domains | Matters, billing, communications, forms and website are distinct domains joined through the matter |
| Provider abstraction | AI, SMS and payments each support more than one provider |
| Configurable workflows | Stages, automation and custom fields are set by the firm |
| Permission-driven access | Roles, granular permissions and tab-level control |
| Human-in-the-loop AI | Draft-first agents with an approval inbox |
| Reusable document structures | One Smart Form structure for manual and AI-assisted creation |
| Workflow automation | Stage changes and schedules trigger tasks, forms, messages and reminders |
| Configurable metadata | Custom fields per practice area, fed by Smart Forms |
| Central platform administration | A Super Admin layer above tenants |
| Optional AI | Core features work without it |
| Provider-independent communication | Platform or firm-owned messaging credentials |
| Integrated billing | Invoices, payments and balances attached to matters |
Five Control Planes
Across these principles, a multi-tenant legal SaaS has to manage five control planes. Matterly’s feature set touches all five, which is why it reads as a platform and not as a bundle of tools.
| Control plane | Question it answers | Where it appears |
|---|---|---|
| Tenancy | Who is isolated from whom? | Firm workspaces, Super Admin, suspension |
| Permissions | Who can do what? | Roles, granular and tab permissions, portal scope |
| Workflow | What happens next? | Stages, automation, Smart Forms, reminders |
| AI spend | Who pays for generation? | AI credits, daily AI Staff budget |
| External providers | Which services carry payments, messages and AI? | Gateways, SMS providers, SMTP, AI provider |
A weakness in one plane tends to surface in the others. A tenancy flaw undermines permissions, and an unmetered AI plane undermines the commercial model. Judging a legal SaaS foundation therefore means testing all five together.
Implementation Considerations
- Provider keys and configuration. AI features need an OpenAI or OpenAI-compatible key. Payment, SMS and WhatsApp features work only once their providers are set up.
- Regional availability. Payment gateways and messaging services differ by country and provider. Confirm coverage before launch.
- Human review. Qualified staff should review AI output, including forms, website copy, translations and drafts, before it is used.
- Operator responsibility. Self-hosting leaves the operator in charge of the operational items listed under Deployment and Security.
- Tenant isolation testing. Verify isolation against the actual implementation before hosting real firm data.
- Regulatory fit. Legal organisations should assess their own regulatory, privacy and security requirements.
- Not legal advice. Matterly’s features should not be read as legal advice functionality.
- Source validation. Check infrastructure and security behavior against the source code before making implementation-level claims to customers.
- Licensing. Read the current licensing terms on the product listing for your intended use, including resale.
Conclusion
The architectural value of Matterly does not lie in any one feature. A form builder, an invoicing module or a Kanban board can each be found elsewhere. The value lies in how they connect.
Four platform properties are layered over that chain. Multi-tenancy lets one installation serve many firms. Optional AI adds assistance without becoming a dependency. Metering through AI and SMS credits keeps costs attributable. Configurable workflows and subscription management let each firm shape its practice and let the operator build a business around it. Together they set an AI legal practice management SaaS built as a platform apart from one assembled from disconnected tools. For wider context, see legal technology.
Matterly is potentially valuable as a foundation for legal SaaS operators and for firms seeking a unified practice-management environment. Production adopters should validate tenant isolation, security posture, regional provider availability and licensing requirements against their own needs.
Evaluate Matterly before you decide
If you are evaluating a legal SaaS foundation, read Matterly’s current feature set, technical requirements and licensing terms before deciding whether it suits your deployment model. Confirm provider availability for your region and which parts of this AI legal practice management SaaS fit your practice, then weigh them against your own requirements.
Frequently Asked Questions
1. What is Matterly AI legal practice management SaaS?
Matterly is a self-hosted legal practice management platform that supports many independent firms from one installation. It runs on Laravel 12, React 19, TypeScript, Inertia.js and Tailwind CSS. One system covers matters, intake, Smart Forms, e-signatures, billing, messaging, firm websites, booking, client portals, optional AI tools and SaaS administration. It is positioned as a legal SaaS foundation.
2. Is Matterly a multi-tenant legal SaaS platform?
Yes. A single installation can host several independent firms beneath a Super Admin layer. Each firm keeps its own staff, clients, matters, billing, website, settings and AI usage. The Super Admin can impersonate a firm to help with support or suspend it while other tenants keep running. Isolation should still be tested against the implementation before production use.
3. How does Matterly Smart Forms work?
Smart Forms are documents that hold legal text, input fields and a signature in one place. Staff build a form by hand or draft it with AI, then send it through a secure link. The client fills the fields inline and signs. The signed PDF is filed under the matter, and staff can review the answers or ask for corrections.
4. Can AI generate legal Smart Forms?
Yes. Staff describe the document in everyday language and AI proposes wording and fields. Staff then review and edit the draft in the visual editor before any client sees it. This needs a configured OpenAI or OpenAI-compatible provider. The output is a draft, not legal advice, and qualified staff should check it before use.
5. Can clients sign Smart Forms online?
Yes. The Signature field lets clients sign electronically inside the form, from a desktop or a phone, after opening their secure link. They receive a signed PDF receipt after submitting. Whether an electronic signature is enough for a particular document depends on the law that applies, so each firm should confirm its own requirements.
6. Can completed Smart Forms generate signed PDFs?
Yes. Once a Smart Form is completed and signed, the platform produces a signed PDF. The client gets a signed PDF receipt, and the submission is filed automatically under the matter. Staff can then review it, and approved answers can be mapped into the matter’s custom fields.
7. Can staff request corrections to individual form fields?
Yes. Staff can flag specific wrong answers and ask for a correction. The client sees only those fields and updates them, and a changes view compares the before and after values for the next review. Nobody has to resubmit the whole form, which keeps the review loop short.
8. Does Matterly include legal matter management?
Yes. The matter workspace is the operational core of the platform. It links the client, tasks, deadlines, forms, billing, communications, activity, custom fields, practice area and stage. Matters move across Kanban boards, and staff can work from one matter record instead of hopping between separate systems.
9. Does Matterly support custom case stages?
Yes. Stages can be given colors and defined separately for each practice area. A stage can also carry a client-facing name that differs from its internal name. Staff drag matters between stages on desktop and use a Move to action on mobile, while clients follow a progress indicator in the portal.
10. Can matter stages trigger automated actions?
Yes. Moving a matter into a stage can create tasks, deliver Smart Forms, send client messages and start other follow-up. Which actions run depends on how the firm sets up each stage. The Kanban board therefore works as an operational workflow, not only a visual tracker.
11. Does Matterly include a client portal?
Yes. Clients activate an account and can see their matters, progress, invoices, balance and relevant information, and they can communicate through the supported channels. The portal is a controlled view rather than a copy of the staff dashboard, and it shows client-facing stage names in plain language.
12. Does Matterly support online consultation booking?
Yes. Visitors pick a consultation and book it, then pay online or choose to pay at the office. The flow includes a payment hold during checkout, a message when a payment fails, SMS consent capture and matter lookup for returning clients. The baseline names no particular calendar provider.
13. Which payment gateways are supported?
Stripe, Razorpay, PayPal and LigdiCash are supported, plus a pay-at-office option. Webhook-based payment workflows are included, and recurring SaaS plan billing runs through Stripe. Availability depends on configuration and on regional and provider coverage, so check that your chosen gateway works in your country.
14. Does Matterly support SMS and WhatsApp communication?
Yes. SMS runs through Twilio, MSG91 and Africa’s Talking, with two-way messaging, delivery status, STOP and START handling, quiet hours and message logs. WhatsApp workflows are supported where the integration is set up. Both rely on provider configuration and are not active by default.
15. Which AI providers are supported?
OpenAI and OpenAI-compatible APIs are supported, and AI features need a configured provider key. OpenAI-compatible support suggests a provider abstraction, but support for arbitrary models is not claimed. The verified baseline names no other providers, so test any compatible service before relying on it.
16. Can Matterly operate without AI?
Yes. Matter management, clients, tasks, calendar, billing, manual Smart Form building, the client portal, websites, communications and permissions all work without AI. Only the AI features, such as Smart Form generation, the website chatbot and AI Staff, need a configured AI provider.
17. Does Matterly support multilingual websites?
Yes. Firm website pages can be multilingual, and the baseline includes multilingual indexing and hreflang. The interface comes in English, French, Spanish and German, with a partial Hindi starter translation. The Super Admin can add languages, and AI translation is available, with staff review recommended.
18. Does Matterly support RTL Smart Forms?
Yes. Smart Forms handle right-to-left text, including Arabic, Hebrew and Urdu. This matters for document-style forms because text direction affects text flow, field placement, punctuation and signature areas, not only alignment. A form that reads naturally in the client’s language is easier to complete.
19. Is Matterly self-hosted?
Yes. It is installed on the operator’s own server through a browser-based installer. The operator controls the deployment and also carries responsibility for infrastructure, updates, backups, security hardening, monitoring and maintenance. Self-hosting is not automatically safer than a managed service.
20. What technology stack does Matterly use?
Matterly is built with Laravel 12, React 19, TypeScript, Inertia.js and Tailwind CSS, and runs on PHP 8.2+ with MySQL 8+ or MariaDB 10.6+. The verified baseline lists no other infrastructure components, so confirm anything beyond this stack with the source code or the seller.
21. Can Matterly be used as a white-label legal SaaS?
It is positioned as a self-hosted, white-label legal SaaS foundation. It supports multiple firms, per-firm branding and websites, custom domains or subdomains, subscription plans, feature flags, coupons, and AI and SMS credits. Operators should read the current listing and licensing terms for their intended use.
22. How does Matterly manage AI credits?
Each firm holds an AI wallet. Requests draw credits down, usage is tracked, and the platform owner manages and monitors credits at platform level. AI Staff agents also run under a daily budget. The baseline does not say how credits are calculated, so no pricing formula should be assumed.
23. How do AI Staff agents work?
AI Staff cover four categories: review replies, client updates, daily task planning and revenue notes. They follow a draft-first model. Each agent writes a draft that waits in an approval inbox, and a person reviews and sends it. They are administrative assistants, not autonomous agents, and they do not give legal advice.
24. What server requirements does Matterly have?
Matterly needs PHP 8.2+, MySQL 8+ or MariaDB 10.6+, a public document root, writable directories and one cron job for scheduled functions. The browser-based installer runs requirement checks before setup continues. Confirm the full, current requirements on the product listing before deploying.
