Legal Services LLM Classification Intake Automation

From First Contact to a Smarter Intake Experience

Making legal intake easier to navigate. How Xmartlabs helped Robinson & Henry improve how potential clients enter the system and get connected to the right attorney.

Client

Robinson & Henry

Industry

Legal Services

Service

Custom Software + AI Integration

Focus

Legal intake & scheduling

Robinson & Henry Schedule Consultation

Context

Making legal intake easier to navigate

Reaching out to a law firm often starts with uncertainty. People don’t always know what kind of legal help they need, what the process looks like, or even what happens after that first interaction. For firms, handling that uncertainty at scale, while keeping a high-quality, human experience, is not trivial.

At Robinson & Henry, improving accessibility meant looking closely at this first step: how potential clients enter the system, and how they get connected to the right attorney.

Challenge

Where the friction was

The firm already had a working intake setup, combining a contact center with tools like forms. It supported their operations, but as demand grew, some limitations became more evident.

01

Rigid entry point

Clients were asked to fit their situation into predefined categories, even when they weren’t sure what they needed.

02

Multiple handoffs

Intake relied on multiple handoffs, from contact center calls to manual attorney review and follow-up coordination. Calls averaged around 15 minutes and still required additional steps afterward.

03

Unclear next steps

Key details like availability, pricing, or next steps were not always clear upfront. The process worked, but it required time, coordination, and interpretation at every step.

Robinson & Henry intake experience

Approach

A more natural way to start the conversation

Working together with Robinson & Henry, the focus was not to replace their process, but to improve how it begins. The result is their Schedule Consultation experience, an intake flow built around a simple idea: let people explain their situation in their own words.

Natural-language intake

Phase 1 · Intake reimagined

Let people describe their case in their own words

Instead of navigating forms, users describe what’s going on. LLMs (via OpenAI) process the input, the system classifies the case into the appropriate legal category, and users are guided to the right type of attorney. This reduces the need for guesswork at the very first step.

How it works

Process input

Classify case

Route to attorney

Connected workflow

Phase 2 · End-to-end flow

Connecting the full intake flow

That initial interaction is only one part of the experience. The solution integrates directly with Calendly for consultation booking, LawPay for payments, and Salesforce for case tracking and follow-up, so once a client starts the process, they move forward without unnecessary interruptions or handoffs.

Behind the scenes

Background processing (Redis, BullMQ)

PostgreSQL for data storage

App layer built with Next.js

Practical AI

A practical approach to AI

A key part of the work was defining how to use AI in a way that was both effective and realistic. Given limited training data, the approach focused on:

LLMs for interpretation

Using LLMs for classification and input interpretation.

Prompting, not training

Relying on prompting and orchestration (LangChain / LangGraph) instead of custom model training.

Latency & cost control

Keeping latency and operational costs under control.

This made it possible to introduce AI meaningfully, without overcomplicating the system.

Results

What changed

The improvements are visible both for clients and for the internal team — faster first interactions, fewer manual routing steps, and a modular foundation that supports future expansion without re-architecting.

50%

Shorter calls

Contact center interactions dropped from ~15 minutes to roughly half that time.

Less manual work

Cases no longer need to be reviewed and routed entirely by hand.

Smoother intake

Prospects can now explain their case in their own words and be matched directly with the right attorney.

Scalable foundation

The modular design (LLM + orchestration + APIs) allows future expansion without re-architecting.

Better-connected systems

Intake, scheduling, payments, and CRM tracking now operate as part of a single experience.

This project focused on a specific moment: the beginning of the client journey. By improving how that moment works, reducing the need for manual routing and connecting key steps in the intake process, Robinson & Henry now handles incoming demand with less friction, while maintaining the human element that defines their practice.

“Xmartlabs is highly skilled, professional, and committed to delivering quality results. Their approachable and friendly demeanor made development stress-free. I recommend them without hesitation.”

Bill Henry

Bill Henry

Robinson & Henry