AAAi Chat Book
AI guidance for the American Arbitration Association.
- launched
- Jan 2025
- editions live
- 4
- of queries
- 1000s
- client through
- 2028

Overview
People navigating arbitration, mediation, or other forms of dispute resolution often face a steep learning curve — unfamiliar rules, complex procedures, and legal terminology that can be hard to parse without professional help. The American Arbitration Association (AAA) wanted to change that by making its authoritative content accessible to everyone, not just legal professionals.
We built AAAi Chat Book as a free, public tool that lets anyone ask questions in plain language and get verified, source-backed answers. Whether you're a self-representing party preparing for an arbitration hearing, a small business owner dealing with a contract dispute, or a legal professional looking up specific rules, the Chat Book draws from AAA's own publications — including Case Preparation and Presentation by Jay E. Grenig and Rocco M. Scanza, AAA-ICDR rules and standards, and specialised editions for labor arbitration and non-attorney users.
What makes it different
Our architecture was deliberately designed to minimise hallucinations. Answers are composed only from the passages the system retrieves, and citations are assigned by the retrieval step, not by the model. We implemented a semantic RAG (Retrieval-Augmented Generation) pipeline that ranks, reranks, and retrieves document fragments by multiple criteria, returning either a precise, source-backed result or an honest "no answer" when nothing in the corpus supports one. Every answer shows exactly which source it draws from, so users can verify independently.
What we delivered
The system provides direct source links, citations, and document previews for every answer. Answers are composed only from retrieved passages, and every citation comes from retrieval, not from the model. When nothing in the corpus supports an answer, the system refuses with "no answer" rather than fabricating a response. Multiple Chat Book editions serve different audiences — case preparation, non-attorney guidance, labor arbitration and mediation practice — each tailored to its users' needs.
The result
AAA now offers a free, publicly accessible AI tool that makes dispute resolution guidance available to everyone — not just those who can afford legal counsel. The Chat Book handles thousands of queries, helping people understand arbitration procedures, prepare their cases, and navigate the ADR process with confidence. It either returns a verified answer or transparently acknowledges the lack of data — a level of honesty that builds the trust essential for a tool people rely on for legal guidance.
At a glance
- Industry
- LegalTech
- Delivery
- AI/ML · Web · API
- Stack
- TypeScript
- Semantic RAG
- AI SDK
- React
- Topics
- LegalTech · AI · Search · RAG
- Timeline
- 2023–2028
In their words
“A breakthrough in making arbitration more efficient, accessible and fair.”
Related articles
- Legal
How We Built AI-Powered Chat Book Tools for the American Arbitration Association
The story behind the AAA's AI-powered Chat Book suite — from architecture decisions to source grounding, and what we learned about building AI that arbitration practitioners trust. - AI engineering
AI Consultant vs Agency, Big 4, Fractional CTO or a Small Senior Team: An Honest Comparison
Who should advise on or build your AI project: an independent consultant, an agency, a Big 4 firm, a fractional CTO, your own hire or a small senior team. Market rates as of October 2026, a comparison table, and when not to hire us. - AI engineering
How Much Does AI Development Cost in 2026? Audit, 5-Day Prototype and $200/h, Against Market Benchmarks
What custom AI development costs in 2026, stage by stage: a $5–10k fixed audit, a $20k five-day prototype and $200/h after that, compared with dated UK and US market rates. Plus fixed price vs time and materials, cost per answer, and a 90-day budget.


