Engagement Letter Drafting Agent · Use case
How can a law, accounting or consulting firm use an AI agent to draft engagement letters on its own infrastructure?
An engagement letter drafting agent drafts the letter from the firm's approved terms and the scope, fees and team the partner has set, and marks every place where the scope or a requested change departs from the approved wording. The partner reviews, signs and sends. The agent accepts no client, agrees no fee and sends nothing. It runs on the firm's own infrastructure.
Every engagement needs a letter.
Each is drafted from the last.
- ✗ 01Letters are copied from the last similar engagement, and old wording, wrong terms or a missing limit on liability travel with them.
- ✗ 02The terms the firm approved sit in one place and the scope of the engagement in another, and checking one against the other is left to the partner.
Where it stops for a person
Always. The partner reviews, signs and sends the letter.Always. Fees, scope, limits of liability and whether the firm takes the engagement are the partner's decisions. The agent never decides that a departure from the approved terms is acceptable.
It drafts the letter.
The partner signs it.
Step · 01
The partner or engagement manager uploads the scope note, with the fees, the team and the client's details.
Step · 02
The agent reads the firm's approved terms for that kind of engagement.
Step · 03
It drafts the letter from the approved wording and fills in the scope, fees, team and dates from the note.
Step · 04
It marks every place where the scope or a requested change departs from the approved terms, and names the term.
Step · 05
The partner edits, approves, signs and sends the letter.
Reads from · connectors
The scope note, as an uploaded file, and the firm's approved terms, templates and clause library, indexed as a knowledge source with each reader's permissions kept.
Acts through · MCP servers
Nothing. It sends no letter, agrees no fee and accepts no client. It writes to no practice management system.
What you set up
- Index the approved engagement terms and templates, by service line, as a knowledge source.
- Give the agent your scope note form.
- Build the agent in Katonic Studio and test it on engagements your partners have already signed.
- Publish it to partners and engagement managers.
Where it runs
The agent, its knowledge store and the model run on infrastructure you own or control. You choose the model and serve it yourself, so questions, documents and answers are not sent to a public AI service. Engagement terms and client details stay inside the firm.
How you get it today
This page is labelled Use case: the agent is built in Katonic Studio. You describe the job, connect what is named above, add test questions and release the agent once it passes them. When the Catalog holds a tested template for this job, this page will carry the Template label and show how to take it.
Does it send the letter?
No. The partner reviews, signs and sends it.
Can it change the terms?
It drafts from the approved terms and marks departures. Whether a departure is acceptable is the partner's decision.
Does it work for every service line?
It uses the approved terms you index for each service line, and says when it holds none for a kind of engagement.