Jennifer Spykerman didn’t just dust off her old dev skills—she rebuilt them using AI. After two decades in leadership, managing multi-million dollar federal programs and complex engineering teams, she opened up a Claude chat and started coding again.

Her mission: automate the mess of federal RFP responses with something smarter than copy-paste.

In this episode, the DefenseLogic AI CEO and founder shares how she shipped a working MVP that flags compliance gaps, tracks requirements, and speaks fluent bureaucrat, all using AI tooling and a custom-built stack.

If your backlog is full of repetitive tasks that no one wants to do, this episode will shift how you think about using AI to get unblocked and ship smarter.

In this episode, you’ll learn:

1. How to use agents to get leverage in locked-down environments
2. What building with Claude and open tools actually feels like solo
3. Where today’s AI tools break and why it matters for platform teams

Things to listen for:

(00:00) Meet Jennifer Spykerman

(02:04) Re-entering code after 20 years away

(05:03) What led her to start coding again

(06:40) The real cost of government proposals

(10:40) Why she built it herself through vibe coding

(15:01) Her AI tool stack and why it works

(19:32) Debugging Claude like a junior engineer

(22:12) Inside the architecture of her AI tool

(28:22) Mastering prompt engineering for consistency

(37:23) Handling security risks for government use

(42:30) How she cracked FedRAMP and NIST with AI’s help

(51:03) Why replacing dev teams is a bad bet