A 90-day AI plan that earns its keep.
Your team has heard "AI" for two years. You've signed up for ChatGPT Teams. Someone tried Notion AI. The marketing team has a Zapier-and-prompt thing that nobody else uses. Nothing has actually changed how you operate. Meanwhile the noise gets louder, your competitors might be ahead, and your board is asking pointed questions about AI strategy.
- 01 Opportunity scan
Every workflow audited for AI fit. Scored on impact × effort × maturity. The top 5 surface. The bottom 30 get a one-line dismissal you can quote.
- 02 Buy / build / ignore matrix
What you should buy off-the-shelf, what's worth building, what to actively ignore. Each call backed by cost, switching risk, and 12-month outlook.
- 03 One shipped automation
I build the first one with your team in week 3–4. Proves the approach, trains the team, gives the board a tangible win.
- 04 12-month roadmap
Sequenced, with budget bands and dependency notes. Defensible at a board meeting; flexible enough to survive the next model release.
- 05 Risk and policy notes
Data handling, vendor selection criteria, the 'what do we do if our LLM provider changes pricing' contingency. Boilerplate-aware, customised.
- 06 Vendor shortlists
Where buying makes sense, you leave with 2–3 named vendors evaluated against your actual workflow — not a category review.
- Teams with 10+ employees and documented workflows
- A founder, COO, or VP with budget and decision authority
- Companies past PMF with operational pain worth automating
- Boards or investors asking about AI strategy
- Solo founders looking for tool recommendations (use ChatGPT and a free afternoon)
- Teams without a single repeatable workflow
- Pre-product startups looking for AI go-to-market advice
- Companies hoping AI replaces a missing strategy
What if the automation you ship doesn't stick?
We pick it together based on adoption likelihood, not novelty. If the team won't use it, we don't ship it. That's the point of week 1 — finding the one where the team is pulling, not pushing.
Will you build with our existing stack?
Yes. The automation lives where your team already works — Slack, Notion, your CRM, your warehouse. Bolt-on tooling that requires a new tab is where AI projects go to die.
What models / vendors do you use?
Whichever fits your data residency, budget, and latency. OpenAI, Anthropic, open-weight via your own infra. Vendor-agnostic by default.
How do we know the strategy is still right in 6 months?
You don't. That's why the roadmap has explicit re-evaluation gates and the buy/build/ignore calls are tied to current market conditions you can re-check.
Can you implement the rest of the roadmap?
Sometimes. Usually I scope a follow-on engagement or recommend an implementation partner. The plan is yours to execute however you want.
Four weeks. A defensible plan. One shipped win. Your board's next question answered before they ask it.