For law firms running Harvey
Automated Legal turns your firm's own Harvey usage data into a personal coaching report for every lawyer, every month, and a scorecard your executive committee can act on. Copilot, ChatGPT, Codex, and Claude covered too. No training sessions. No lost billable hours.
Validated inside large law firms. Run by the founder of one of the first generative AI companies.
Partner, Corporate · Monthly report, with history from February
Monthly activity
You, next to the firm
| Measure | You | Partners |
|---|---|---|
| Active days in July | 18 | 9 median |
| Prompts citing a source | 58% | 71% median |
| Threads with a follow-up | 31% | 19% median |
One of your prompts, scored
"summarize the key risks in this agreement"
Your firm made the Harvey investment, and Copilot, ChatGPT, and Claude are probably in the building too. The invoices arrive every month. What arrives with them is anecdotes: one associate who swears by it, one partner who tried it twice in March, and a renewal decision nobody can support with evidence.
When we instrument a firm's AI stack, the same picture appears. A minority of lawyers become regular users. Most prompts leave out the context, the audience, and the format that make output usable, so the first response disappoints and the lawyer quietly goes back to the old way. The spend stays. The habit doesn't form.
Training was supposed to fix this. But a lunch-and-learn costs a room full of billable hours and evaporates within a week, because it isn't about anyone's actual work.
Who actually uses the tools you pay for?
Are they using them well?
What changed since last month?
154 licensed seats, one month
| Question the Scorecard answers | Example finding |
|---|---|
| Seats never used this month | 92 of 154 |
| Practice group furthest ahead | Corporate |
| Share of prompts missing key context | 63% |
| Month-over-month engaged users | +12% |
Your administrator sends a Harvey usage export or grants read-only API access. Nothing is installed. No one touches your document management system.
Each substantive prompt is scored against a legal prompting framework: context, task, audience, instructions, output format, and verifiability. Firm-level patterns are benchmarked month over month, by role and practice group.
Each lawyer receives a personal report: their own prompts analyzed and rewritten, the patterns worth keeping, and one specific experiment for the next 30 days. Leadership receives the firm Scorecard.
Total lawyer hours required: zero.
Eight pages, personal to each lawyer, generated from their own work. Here is what the four core sections do. All examples below are fictional.
Usage benchmarked against the firm and against peers in the same role. Not a leaderboard. A private mirror: most lawyers have no idea whether they're ahead or behind, and finding out changes behavior on its own.
Every substantive prompt is scored on six elements. The report shows the lawyer's strongest prompt and their leanest one, side by side, so the difference is visible in their own words rather than in a slide.
Real prompts, rewritten with the minimum change that would have made the response usable the first time. One reusable pattern per rewrite, so it transfers to next week's work.
Why it helps: role, stakes, and format let the tool produce something a lawyer can send, not something a lawyer must rebuild.
A single habit to test, chosen from the lawyer's own gap, plus one capability they haven't touched that fits the work they already do. Next month's report shows whether it moved.
Free · for anyone
PDF, fictional lawyer, real format
Free · for your firm
One usage export in, one scorecard back
The program
Priced per lawyer per month · scales with firm size, not our hours
Not ready for monthly? A one-time Legal AI Audit covers the firm Scorecard plus coaching reports for your 25 most active users, with a findings readout. If you continue into the program, the audit fee is credited.
We can make that promise because we measure the result in your own data, and we show you the same numbers we hold ourselves to. No adoption lift, no bill.
Coaching tells you where the gaps are. Some firms then want help closing them. The same team, the same person, hands-on.
Automated Legal is Automated Consulting Group's advisory service for law firms, run by Robbie Allen. Robbie has spent 20+ years building AI companies. He founded Automated Insights, one of the first commercial generative AI companies, which produced millions of automated narratives a week for the Associated Press and Yahoo before it was acquired. He has since co-founded and sold two more companies, written ten technology books, and now advises mid-market companies and law firms on making AI actually work.
The coaching methodology was developed and validated inside large law firms: months of real usage instrumented, thousands of lawyer prompts analyzed, reports in lawyers' hands. This is that system, made available to your firm.
"The leadership facilitation session completely changed how our executive team thinks about AI. We went from fear and skepticism to a shared vision in a single day."
CEO, regional law firm (ACG client)Coaching requires reading prompt text, and prompts can reference matters. So we treat this the way your firm treats any engaged professional: written confidentiality obligations in the engagement agreement, processing through enterprise AI services configured for zero data retention, and identifier scrubbing in every report we deliver. Your data is never used to benchmark other firms without your written consent.
If your firm prefers, we can run a metadata-only engagement: full adoption analytics and scorecards, lighter per-lawyer coaching.
Harvey first. That's where the analysis is deepest: prompt-level scoring, feature utilization, and per-lawyer coaching from the actual usage data. If your firm also runs Copilot, ChatGPT Enterprise, Codex, or Claude, we fold them into the firm Scorecard where the vendor's export supports it, and we tell you exactly what your stack supports before you sign anything.
No sessions, no workshops, no room full of lawyers watching slides. Each lawyer gets a written report about their own work, on their own time. It takes about ten minutes to read. That is the entire ask.
No. We are independent and vendor-neutral. We analyze the tools; we don't sell them. If your usage data says a tool isn't earning its seat cost, the report will say so.
The sample report and the firm Scorecard are free. The monthly program is priced per lawyer and typically runs a fraction of what your firm already pays for the AI licenses themselves. Exact pricing depends on firm size and tool mix; you'll have a number after one conversation.
Download the sample report in one click, or send one export and get your firm's Scorecard. Either way, you'll know more about your AI spend than you do today.