Law firm AI training: a step-by-step guide

A law firm’s inbox is always full of the same questions: "How long does a divorce case take?", "I couldn’t collect my severance pay, what should I do?", "How much is the consultation fee?", "I filed a traffic accident report, can we open a case?". Family, labor, criminal, commercial, enforcement, real estate: whatever the area, the same twenty questions come in dozens of times a day. While the secretary tries to keep up with the phone and WhatsApp, a prospective client who writes after hours waits until the next day. Every consultation request answered late goes to another firm.

inTusell takes over this load. But to do that, it first needs to learn your firm, your practice areas, and your consultation routine. This article explains how a law firm trains the inTusell AI from scratch, step by step. This is the law chapter of our sector-based "how to train your AI" series, and it is the series' main article.

Quick answer

Why does training matter?

inTusell is not a simple chatbot; it is an AI sales assistant and CRM that answers messages on WhatsApp, Instagram and web chat, books appointments, follows each prospective client through the pipeline and behaves like an experienced firm secretary. No good secretary starts on day one without knowing the firm: they need to know which areas you practice in, your consultation routine, and which questions to leave to the attorney. Training teaches the AI to do two things well. First, give the right information: your practice areas, the consultation process, the required documents. Second, stop at the right boundary: leave the legal interpretation to the attorney, and never promise a case outcome.

An untrained assistant either gives overly generic answers or guesses on a topic it doesn’t know. In law, guessing is dangerous. Wrong deadline information leads to a loss of rights, and a boundary-crossing line like "you will definitely win this case" both misleads the client and violates the rules of professional conduct. A well-trained assistant stops when it doesn’t know and directs the client to the attorney.

Who is it for?

This guide is for law firms whose inbox is dominated by consultation requests and who answer the same questions over and over again:

• Firms working in high-consultation areas such as family, labor, and consumer law
• Offices handling criminal, commercial, enforcement-bankruptcy, and real estate cases
• Firms receiving heavy applications in traffic accident, compensation, and insurance disputes
• Teams getting dozens of preliminary inquiries and consultation requests a day from Instagram and WhatsApp
• Small and mid-sized offices that also want to respond quickly to messages arriving after hours and on weekends

You don’t need a technical team. All of the steps below are completed from sections in the panel, without writing code.

The structure of training: 5 core steps + 2 continuous steps

Splitting the training into two groups makes your job easier. The first 5 core steps get the assistant up and running; this is the one-time part of the setup. The 2 continuous steps that follow (reviewing replies and turning fixes into rules, and opening channels) are the ongoing part that you sharpen as you use the assistant.

1. Upload the knowledge base

The foundation of training is the knowledge base (RAG). In the panel’s Knowledge Base section, you upload all of your firm’s textual information. Each time the AI generates a response, it automatically searches these sources and uses only the information written there.

Supported source types:

Every file you upload is automatically chunked and made searchable with pgvector. When you upload a "documents required in labor cases" document, the AI finds the right paragraph and responds when a client asks "I was dismissed, what should I bring?".

What a firm must definitely upload: practice areas and case types, how the consultation process works, the documents to bring to the first meeting, working hours, team and attorney information, the communication policy, and the most frequently asked questions. An important limit: the knowledge base is not for legal-opinion-containing text or person-specific case information; it is used in every conversation. The AI uses the information here to answer "which areas do you practice in, how does the process work, which documents are needed" questions; it leaves legal questions like "am I in the right in my case, how much compensation will I get?" to the attorney.

2. Set up the consultation appointment system and working hours

The knowledge base covers "what you offer"; the appointment system lets the assistant give the client an actual time. In a clinic this step was the examination appointment; in a law firm the equivalent is the consultation appointment. In the panel you define three things:

Once these three are defined, the AI no longer invents times; it suggests a consultation appointment based on your working hours and full slots. The system blocks two appointments at the same time (double-booking); if you add your attorneys as staff, each one has their own hours and the check runs per attorney. If you write a preparation note on the consultation type (documents to bring), it appears in the 1-day reminder. If Google Calendar is connected, each new appointment is also copied one-way to the team’s calendar; an event added in Google does not block the slot in inTusell.

Alongside the consultation appointment, you can add a second routing capability: intake routing. The AI asks which practice area the matter falls under, tells the client which documents to bring and routes them to the right consultation type, but it does not interpret the matter or assess the merits. We’ve gathered all the details of the appointment engine (durations, conflict prevention, reminders, and intake routing) in a separate article: consultation appointment and intake routing automation for law firms. You can also see the service and appointment capabilities on the solutions page.

3. Set the persona and tone

You taught the AI what to say in the first two steps; now you’ll define how it will say it. There are two settings in the panel:

• Assistant name (ai_persona_name): The name it will introduce itself with to the client. Most firms choose a reassuring, formal name consistent with the office identity.
• Tone (ai_tone): For law, the default tone is professional; respectful, measured, reassuring, and mindful of confidentiality. Emoji use is off and sales pressure is kept low.

The law firm sub-type is deliberately set up more cautiously than other sectors. Even the ready greeting points the client in the right direction: "Hello, welcome. I can help with your meeting request and how the firm works. What would you like to discuss?" The ready instructions also keep the AI from using phrases such as "the best", "expert", success rates or examples of won cases, under the Attorneys’ Act and the advertising ban. This line brings the firm’s tone and the requirements of professional rules together in one place.

A multilingual note: the AI replies in whatever language the client writes in. For firms with foreign client traffic, or those handling international citizenship and residence matters, messages arriving in English, Arabic, or German need nothing extra from you; the assistant detects the language and replies in it while keeping the legal tone. Fixed system messages such as reminders go in Turkish to Turkish speakers and in English to everyone else.

4. Define the response rules and legal limits

The persona defines "who it is," and the response rules define "how it will behave." In law, this step does two jobs at once: conveying your firm’s consultation discipline and drawing the legal limits. With the AI Manager Chat, you add rules by writing in plain language, just as you would instruct a teammate.

Typical rules for law firms:

• "Before the consultation, ask which area the matter falls under and say which documents to bring, but do not ask for documents and do not give a legal interpretation."
• "When asked about the consultation fee, do not give a precise amount; say the fee will be determined in the meeting and direct them to a consultation." (Can be changed according to your firm policy.)
• "Do not predict case outcomes; do not use phrases like 'you will definitely win' or 'you will definitely get compensation.'"
• "If a legal assessment or an urgent legal situation is involved, direct the client to the attorney; for urgent criminal matters such as custody/arrest, escalate to the live team immediately."

The most common mistake in law is misleading the client, and running afoul of the rules of professional conduct, with a single boundary-crossing sentence. In inTusell, this boundary is not left to the rules you write alone. The main control is the instruction set that comes ready for law firms: legal opinions, chances of winning, sentence or compensation estimates and outcome promises are forbidden; if a deadline risk is sensed, the AI does no deadline calculation and hands the conversation to the team. On top of that, an output filter (a pattern scan plus a second-model check) scans for misleading legal promises such as "you’ll definitely win the case" or "guaranteed compensation" and flags them. The filter is not a wall that blocks everything; it is a second safety net that backs up the instructions. Legal and sensitive topics reach your team in every mode.

To compare two AI tones, you can use the A/B test, which is in beta: for example, a more formal and a friendlier tone run side by side, each on half of the replies, and the experiment card shows which one leads to more wins. Today only the tone changes; the content and the rules stay the same.

5. Teach from past conversations

This is the step that moves training from "good" to "specific to your firm." In the panel, you upload audio recordings of your past client conversations (MP3, MP4, WAV, M4A) and mark each one as Won or Lost, for example, whether the client came to the consultation and an engagement was formed, or was lost.

The system uses these recordings in two ways:

This way, the AI learns from your won calls how your best coordinator put the client at ease and moved someone who said "let me think about it" toward a consultation. The KVKK side is protected: uploading requires explicit consent, and phone, email, ID and card numbers are masked in the transcript before processing. Because client confidentiality is involved, this protection is especially important in law; the AI learns the coordination pattern from the recordings and does not draw a legal opinion about the case.

This step is not mandatory, but don’t skip it. The knowledge base teaches the AI "what" it knows; past conversations teach "how you coordinate." When the two are combined, the assistant truly resembles an experienced consultation coordinator.

6. Review and turn fixes into rules (continuous step)

The first five steps get the AI up and running. What follows is the continuous part that sharpens it over time. You check replies on the AI Test screen and in Conversations. Say the AI explained a document list incompletely, or referred to a practice area by the wrong name: you add the right fact to the knowledge base or write a rule in AI Rules. You can switch the ready rule packs in the Behavior Library on or off, and write a plain instruction such as "if asked about fees, explain the tariff and the meeting" in the AI Admin Chat. The process names and standard information sentences you use build up the same way.

The Label Manager is a separate quality record: customer satisfaction replies, the outcome of handed-off conversations, the won/lost classification and labels a team member adds by hand land there. You approve, reject or edit them; these corrections are kept as a record, but they do not change how the AI writes replies today. What changes behavior is the knowledge base and the rules. If you want to see every reply before it goes out, use Full approval mode or the conversation’s Approval mode.

In the first weeks, we recommend a short daily look at the replies. Review is the "live" part of training.

7. Open the channels (continuous step)

Once training is ready, you put the assistant in front of clients. inTusell gathers all channels in a single inbox: Instagram DM, Instagram comments, WhatsApp, Telegram, and Web Chat.

Channel-opening methods:

• Instagram: connects through Meta login, where you pick your Page and Instagram account; a Professional account linked to a Facebook Page is required.
• WhatsApp: connects by scanning a QR code or with an 8-character linking code sent for your phone number; your number stays the same and no Meta Business approval is required. The official WhatsApp line (Cloud API) opens after Meta approval.

The practical path is this: open the one-to-one messaging channels first, WhatsApp and Instagram DM, and automate the consultation traffic; DM replies are already on once Instagram is connected. Instagram comment automation is live too, and the "Instagram comments" switch on the Instagram card is on by default. If you like, switch it off, answer the "can I get information?" comments under your posts by hand at first, and switch it on when you are ready. If you connected Instagram before September 24, 2026, you need to reconnect it once for comments.

There are three account-wide modes (My business › Instructions › AI Settings › Advanced AI Settings): Intern (default; sends replies itself, asks you before a cancellation or refund), Senior (also handles cancellations and refunds itself) and Full approval (every reply on WhatsApp, Instagram and Telegram waits for your approval). In every mode the AI handles consultation booking and information traffic; conversations that need a legal assessment go to the attorney, and the handoff lands on the Decisions page. You can also put a single conversation in Approval, Shadow or Paused mode. For the details of Instagram and WhatsApp automation: Instagram and WhatsApp consultation automation for law firms.

How long does training take?

How long it takes depends mostly on whether your documents are ready. The steps of a working setup:

1. Uploading the first files to the knowledge base
2. Consultation appointment types and working hours
3. Persona, tone, and first response rules
4. Connecting channels

But there is no moment when "training is done." Over the first weeks, as you add the gaps you see to the knowledge base and AI Rules, the assistant gets closer to how your firm works. Uploading past conversation recordings and tone tests are ongoing improvements too. Setup is fast; mastery is continuous.

What it isn’t

Putting inTusell in the right category matters, because the wrong expectation leads to the wrong setup.

• It is not an attorney or a legal advisor. It does not give legal opinions, predict case outcomes, or recommend strategy. It directs legal questions to the attorney and the firm.
• It is not a legal process/case file management system (a court e-filing client or case-tracking software). It does not run case files; it is a consultation appointment, communication, and pre-routing layer.
• It is not a marketing tool that promises outcomes. Phrases like "you will definitely win / guaranteed compensation" violate the rules of professional conduct and are forbidden in the ready instructions; the output filter flags them as well.
• It is not an intermediary that replaces the firm. The attorney always provides the service and forms the engagement; inTusell is not a party to payment.

In short: not a bot that gives canned answers, but an AI sales assistant and CRM that represents your firm to the extent you train it, gives no legal advice and leaves the legal question to the attorney.

Frequently asked questions

How long does law firm AI training take?

A working setup comes down to these steps: uploading the knowledge base, consultation appointment types and working hours, persona and tone settings, and a few response rules. How long it takes depends mostly on how ready your documents are; the real refinement comes in the first weeks, as you add what you notice in AI Test and in real conversations to the knowledge base and the AI rules. Training is not a one-time job; it is continuous.

Does the AI give clients legal advice or an opinion?

No. With the instructions that come ready for law firms, the AI gives no legal opinions, predicts no case outcomes and recommends no strategy; legal and sensitive topics go to the attorney and the team in every mode. These instructions are the primary safeguard; in addition, an output filter (a pattern scan plus a second-model check) scans for misleading legal promises such as "you will definitely win the case" and flags them.

Do I need technical knowledge to train the AI?

No. All training is done from the panel; no code, API key, or developer is required. You upload PDF/Excel files to the knowledge base, enter consultation types and working hours, and write response rules in plain language. You connect Instagram through Meta login and WhatsApp with a QR code or a linking code.

Are client data and case file details safe under KVKK?

Data is stored encrypted and each business’s data is kept separate from the others. You download your account data with "Download My Data (KVKK)" and manage the KVKK notice your clients receive, with your own privacy link, under "Legal notices". The AI does not save high-sensitivity fields such as the practice area without your confirmation, and consent is recorded per channel and purpose, with its time. The AI produces no opinion on a case and does not ask for documents or case details in the chat on its own.

Does the AI quote a consultation fee or attorney fee?

In the instructions that come ready for law firms, the AI states no amount and does not negotiate fees: it says fees are set by the Attorneys’ Minimum Fee Tariff and a written engagement agreement, and that the exact figure is given in the meeting. You can adjust the instruction to your firm’s policy; the AI never invents an amount it isn’t sure of.

Does the AI create the consultation appointment on its own?

Yes. The AI recommends the appropriate consultation type and creates the appointment while accounting for working hours and full slots; double-booking the same time is prevented by the system. In cases requiring a legal assessment or complex situations, it hands the conversation off to the attorney.

Next step

You’ve trained your assistant; next comes using it in daily operations. The next article in the series explains how to run the inTusell you trained on a real workday: incoming consultation requests, calendar, reminders, reducing no-shows, and handing off to the attorney. Continue straight from there: how a law firm uses inTusell. You can also review the capabilities specific to your firm on the law firm solution page.

This series is written separately for each sector. You can read the same setup for tourism in tour agency AI training, for health in clinic AI training, and for insurance (another consultation-heavy area) in insurance AI training.

If you’d like to see it live before starting the setup, use Get a demo and we’ll open your panel together in a short session, or write to hi@intusell.com. For package and quota details, you can check the pricing page, and for other guides, take a look at the all articles list.