How this site works
This site is working software, not a brochure. It’s how I run an AI-engineering and QA practice in the open. You can see the proof, scope your project with Nadine, and get a real plan and price before anyone gets on a call.
#What this site is
I’m Jason Teixeira. I build AI features (chatbots, agents, RAG systems, voice, automation) and the evaluation and testing systems that prove those features actually work. This site is that practice, running in public. The tools on it aren’t mockups. They’re the same kind of software I build for clients, tested the same way, so every page doubles as a live sample of the work.
#How it’s organized
Four surfaces, one purpose. Each answers a different question you might have on the way to deciding whether to work with me.
See the proof
Case studies, public repos, and a live eval you can run in your browser right now. Every claim on the site links back to a receipt.
Scope a project
Scope Studio turns a few questions, or a chat with Nadine, into a priced and itemized plan. This is the main path, and where most people start.
Learn the method
A documentation and guide library on evaluating, testing, and shipping AI. The same method behind everything I build, written out in full.
See the services and automations
What actually gets built: AI features, eval harnesses, test suites, workflow automation, and live demos like the AI front desk voice agent.
#The path through it
However you enter, everything funnels toward one place: a scoped, priced project. The route is short, and nothing on it skips a step or invents a number.
#Scope Studio and Nadine
Scope a project is the primary next step and the fastest way to a real answer. It starts with a handful of quick questions, or a conversation with Nadine if that’s easier. Either way the plan you get back is computed, not generated: every line item comes from a fixed rate card, and the AI’s job is to route your answers into the right line items, never to guess a price. You get an indicative range, itemized by phase.
Nadine is the assistant in the corner of every page. It’s grounded in what I actually build and what this site actually says, so it answers about the real practice instead of free-associating about my career. Ask it a question, or use the mic and talk. Ask whether it’s AI and it tells you the truth. I red-teamed it in public and published the pass/fail results on I red-team my own AI.
Email a plan to yourself and it lands in your inbox, plus a short follow-up from me, because I read every plan Scope Studio produces. No signup, and no call required just to see the number. Want the same idea wired for a real business instead of a portfolio? The AI front desk demo answers, qualifies, and books as a live voice agent.
#After you scope
The Scope Studio number is real, but it’s a range. What happens next mirrors a real engagement. I turn the scope into a firm, itemized proposal and review it myself before it goes out; nothing auto-sends. You accept it and pay a deposit to start, the deposit reserves the work and is credited to the total, and the balance is invoiced on delivery. The work is fixed-scope, so what’s itemized is what gets built, and anything new is quoted before it starts. The full version is documented in How engagements work.
#Why you can trust what you see
The rule behind this whole site is simple: no fake green. A claim only counts if you can check it.
- Every capability links to a real artifact: a public repo, a verbatim test run, or a live screenshot.
- This site runs its own Playwright and axe QA suite on every deploy, over a hundred checks gating each release. It’s held to the same standard it sells.
- Scope Studio computes prices from a fixed rate card, so the number you see traces back to line items you can read.
- A human reviews every proposal before it reaches you. Nothing important auto-sends.