From founder or freelancer
Technical founders, early startup engineers and freelancers who have built products for clients and want to deploy AI inside larger companies.
What already counts
- Owning work end to end with nobody telling you what to do
- Talking to customers, finding what they will pay for and shipping fast
- Wearing many hats: frontend, backend, data and support
- Scoping work against a budget and a deadline
What to add
- Enterprise constraints: security reviews, identity, privacy rules and slow approvals
- Eval discipline and staged rollout instead of ship and see
- Working inside someone else's repo, process and change rules
- Writing updates and handover documents for large organisations
- Feeding field learning back to a product team you do not control
1Map the roleWeeks 1 to 2
See how FDE work at an AI company differs from running your own product or client work, and score yourself honestly.
Show for it: A readiness checklist with proof for each item and a note on which founder stories show agency, discovery and delivery.
2Enterprise constraintsWeeks 3 to 5
Learn what slows big companies down and how to move through it: stakeholders, identity, security reviews and personal data.
Show for it: A stakeholder map and security answers for a bank project, and a Customer Room session at Kaveri Bank scored solid or better on constraints.
3Proof, not only demosWeeks 6 to 8
Add the discipline enterprises expect: evals before launch, guardrails, shadow mode, monitoring and a clean handover.
Show for it: The claims extraction take-home with a shadow-mode report and a rollout plan that has a go or no-go number at every stage.
4In their repo, on their rulesWeeks 9 to 10
Practise working inside code and processes you do not own, and writing for a large organisation.
Show for it: A pull request description and a weekly update that a customer's change board would accept without questions.
5Tell the story and interviewWeeks 11 to 12
Present your company or client work as FDE evidence and prepare for the loop.
Show for it: A one-page summary of your best customer engagement with the problem, what you built, the numbers and what you would do differently, ready for the project deep dive.