Start with the 90-Day AI Program.
One quarter with a Chief Data & AI Officer. Everything else follows from the roadmap.
Start free.
Not sure where you stand? Take the free AI Alpha Ladder: 35 questions, five minutes, no call attached. You get your step on the ladder, the factors holding you there, and a 90-day plan.
Take the free AI Alpha Ladder →The 90-Day AI Program.
Be the person who gets AI running at your firm.
Your Chief Data & AI Officer for one quarter, led by Khurram himself. Month one maps every team and writes the firm-wide roadmap. Month two stands up the shared record and puts the first workflow live. Month three measures it against today and starts the second.
You walk away with the roadmap, a plain-English data plan, a signed AI policy, one workflow in production with a runbook, and the results in numbers the CFO accepts.
You can stop at day 30 and keep everything. If the day-30 roadmap does not find value worth ten times that month’s fee, month one is free.
Read the full detail →- Time
- 90 days, two days a week
- Fee
- Fixed monthly, agreed before work starts
- Terms
- Stop at day 30 if you want, and keep everything delivered
- Led by
- Khurram Tehseen, personally
Two to three firms at a time. Senior people do the work, so the number stays small.
The seat stays, and the build follows the roadmap.
Most firms keep the Chief Data & AI Officer seat on a monthly retainer: a workflow per team, the quarterly re-ranking, and the build work below when the roadmap calls for it. Came here because something specific just happened? Talk to us about the situation directly.
Pilot-to-Production Audit
A pilot that demoed well, cost real money, and has no test set and no named reviewer. Scored go, fix or stop, in writing.
- Time
- 2 to 3 weeks
The Replay
Nobody measured the old way, so nothing can be ranked or defended. One past close run again on your own files gives you the baseline.
- Time
- 4 weeks
Eval Harness Build
A workflow is running and nobody can say how accurate it is. The 20-item test set and the accuracy bar, re-run on every change.
- Time
- 4 to 5 weeks
Exam-Ready AI
An examiner, auditor or LP is going to ask how the model works. Citations, an incident path and the file that answers the published question set.
- Time
- 2 weeks to the dry run, 5 to the file
For private credit managers, the build is usually on the firm’s own record: covenant compliance recomputed from the agreements, borrowing base verification line by line, and the close against the administrator. More on private credit →
Build work runs with a bench of senior engineers across data engineering, AI engineering and model governance, under Khurram’s review.
Every build includes training for the people who run it.
Work that does not fit a door. A board session, a second opinion, an AI section for a fundraise. Ask.