About Jigg.AI

Jigg.AI is an AI-accelerated software engineering studio with a public build journal. AI does much of the hands-on work — coding, prototyping, testing, documentation. An experienced engineer directs it, reviews it, and is accountable for the result.

How the work is done

• The human: defines the problem, sets priorities, makes the engineering decisions, reviews every result, and approves every deployment.
• The AI: writes and changes code, builds prototypes, runs tests, drafts documentation, and iterates.

On an existing system, for example, the engineer maps the architecture, dependencies and risks; AI helps analyze, refactor, migrate and write tests; the engineer then checks the results, catches wrong assumptions, and makes the decisions that matter.

Human-directed isn’t the same as hands-off. Even when AI writes every line, choosing the goal, judging the result and deciding to ship are human work — and that’s the part Jigg.AI doesn’t automate. On client work it’s also where accountability sits: the engineering lead, named below, answers for what’s delivered.

The build journal

The build logs are Jigg.AI’s own projects, documented in public: a website, a support chatbot, an explainer video, a hardened signup endpoint. Each one is tested against a check decided before building, and written up with the mistakes and dead ends left in. They’re the evidence for the studio — you can see the method before you ever get in touch.

Client work is separate, and private by default. Nothing about a client’s project is published unless the client chooses to share it — then only what they approve in writing, and they review it first. How sharing works.

The build logs are written in Jigg.AI’s voice: the “I” in them is the builder persona, not a person. Builds and write-ups are produced using AI tools, under human direction. Results, limitations and verification status are documented for each project.

Why transparency

First-hand, tested, documented proof is exactly the kind of thing that can’t be faked after the fact — “it invented pricing on attempt one, here’s the fix.” Openness is both the differentiator and the reason to trust any of it. Where a build log links to a tool, that link may be an affiliate link; it’s always disclosed and always placed after the proof. Client work is different: tools recommended to clients earn Jigg.AI nothing.

Built with the tools it reviews

This site is itself a build. It’s made with Astro and Claude Code, logged as it was built, and published as the first build log. The whole thing is plain files in a repository — every build a single entry, no database, no lock-in — so the structure is reproducible, and nothing here is hidden behind a black box.