Put AI to work in your company, properly.
I spent nearly twenty years building software and leading engineering teams. Now I help business owners and engineering leaders get real work out of AI. Training on your team's own work, one workflow built and running, and the engineering rollout when you have a development team.
The first call and the written proposal are free. Prices are in the proposal, not on this page.
Can you confirm phase 2 delivery and resend the quote?
The request your front office answers forty times a week, with the details in three different systems.
Two kinds of teams, one way of working.
Whether the work is in the front office or in the codebase, the shape is the same: AI drafts, a person decides, and we measure the difference.
Business owners and operators
Your team already has ChatGPT open in a tab. Nobody has shown them how to use it on the work that fills their week. I run a working session on your own documents and examples, then build the one workflow that costs you the most hours: intake, reports, follow-ups, scheduling. A person checks every output before it goes anywhere.
Engineering leaders
Your developers have AI coding tools and adoption is uneven. I set the standards, run the rollout, measure the before and after, and rebuild the delivery pipeline so AI drafts at every stage and a named engineer approves at every gate. The last time I did this across a 65-person organization, adoption reached 100% in under four weeks.
What I do, and the shape of each.
Six ways to start, from a free session to a four-week engagement. Every one of them ends with your team able to run it without me.
On your company's documents. You leave with three repeatable skills built on your own examples, and a team that knows how to use them.
The first session is on your laptop and costs nothing. After it, a schedule for the first month and a review at day thirty.
A free call and a two-page proposal first. Then one workflow live, with support after go-live. Intake triage, credential tracking, weekly reporting.
Coding, review and test tooling rolled out across the teams, adoption measured weekly, and a before-and-after you can show the board.
Over the agents you already run: evaluations, observability, cost and failure modes. Written findings and an evaluation harness to start from.
For teams without a senior engineering leader.
Three steps, and you can stop after any of them.
No retainers up front and no paid diagnostics. The call and the proposal are free, and the proposal has two sizes on it.
- 30 minutesA call.
One question: which task, done by which person, costs you the most hours every week?
- Within 48 hoursA two-page proposal.
Scope, deliverables, schedule and price, with a small start and the full engagement both on the page.
- Day 30 reviewThe work, with your team.
Built on your tools, handed over with training, reviewed at day thirty.
Numbers from the last engineering organization I led.
From my engineering leadership work before CogniStack, across a 65-person delivery organization. Company names withheld.
Three rules I work by.
They are in every proposal and every statement of work, so you can hold me to them.
- 1Measure first.
A baseline before any recommendation, so the gain is provable and the report is honest.
- 2A person at every gate.
AI drafts, your team decides. Nothing ships, sends or merges without a named person.
- 3You own everything.
You buy your own AI seats, you keep every deliverable, and your data stays in your systems. No patient data, no credentials, no production secrets in a first build.
Firas Taiem
Nearly twenty years in software and more than a decade leading engineering teams, most recently a 65-person delivery organization across the US, Canada and the Middle East. I speak to owner groups about putting AI to work, and I still write code most days.
Based in Canton, Michigan. Remote across the US, on site in Michigan.
Start with a call.
Tell me what your team is trying to get done. I read every message and reply myself, usually the same day.