Engagements with clear scope and measurable outputs.

What we do

Interviews with the team, reading the current stack, mapping data flows. Workshop to align scope for the pilot. We don't write production code in this phase - just throwaway prototypes to validate hypotheses.

Deliverables

Architecture document. Map of current state and target state. Pilot execution plan with stack decisions, identified risks, and cost + timeline estimate. You keep this even if you don't continue with us.

Timeline

2 calendar weeks, starting Monday of the week after contract. Kickoff workshop at the start, final presentation on the last Friday. This is a fixed window.

Risks & limits

Discovery doesn't guarantee the pilot will work. What it guarantees is an informed decision. If at the end the recommendation is not to build, we say so. Client keeps the report all the same.

What we do

We build the minimal data foundation the use case requires (not the universe) and the agent on top. We deploy in small production - real access, real data, real users (internal or pilot client). We monitor metrics during operation.

Deliverables

System running in production with limited scope. Documentation. Your team's access to the repositories. Operational metrics dashboard. Automated agent evals. Everything we built is yours, even if you don't continue with us to Production.

Timeline

4 to 8 weeks depending on the use case complexity. Weekly milestones, demo every 2 weeks. You follow in real time, not just at the end.

Risks & limits

A pilot is a pilot. It can fail in production on cases we didn't test. Rollback is documented. Volume is intentionally limited, so the blast radius of any failure is small and recoverable.

What we do

Pilot expanded to every use case defined in Discovery. We harden infra (Terraform, CI/CD, monitoring). We build evals running in CI. We train your team to operate it. We define a joint SLA.

Deliverables

Full platform in production. Documented repositories with architectural decisions. Operational runbook. Handoff plan with knowledge transfer sessions. Eval suite. Monthly SLA and metrics.

Timeline

3 to 6 months depending on scale. Incremental releases every 2 weeks. Each release has clear scope of what landed and what's left. No big bang.

Risks & limits

Client stack changes during the engagement (ERP swap, M&A, cloud migration) can shift scope. We renegotiate honestly when it happens. We don't hide behind contract addenda.

What we do

Stack decisions, code review, hiring, technical roadmap, critical code review, eval and metrics definition. We work alongside the team running day-to-day operations. Senior brain shared between internal team and strategic questions.

Deliverables

Availability defined by contract (typically 1-2 business days per week). Weekly team meetings. Monthly roadmap review. Code review on critical PRs. Direct channel availability for urgent questions.

Timeline

Monthly. 3-month minimum. After that, cancelable with 30 days notice. No exclusivity clauses. No early-termination fees.

Risks & limits

We're not replacements for an internal team. If the company has nobody to run day-to-day, Fractional CTO alone isn't enough. We discuss this honestly at intake if we identify it.