We build AI features — and the proof they work.
Sage Ideas is a one-operator AI engineering and quality studio. For thirteen years my job was making sure software actually worked — the kind of QA where a missed bug costs real money. Now teams ship AI features on a demo and a prayer, and customers find the breakage first. So I build both halves: the AI feature, and the automated evidence that it does what you claim. This document is the short version — every figure was counted from source, and you can open the front door of the products yourself.
What we do
Two things, tightly coupled, because shipping AI without proof is how good products break in public:
- Build the AI feature — LLM-backed assistants, RAG systems, agents, and the workflow automation around them, wired to real models and real data.
- Prove it works — evaluation harnesses, test automation, and CI gates that make "it works" a measurable claim, not a vibe. Citation coverage, regression suites, eval scores that go red when they should.
The output is always the same shape: a feature that ships, the evidence it behaves, and a system your team can run after I'm gone — your repo, your CI, nothing retained after handoff.
Who you work with
You work with me — Jason Teixeira — directly, start to finish. No account managers, no handoff to a junior. That's possible because the same engineering system that runs my own two live products runs your build. Solo, at production scale, is the whole proof: hiring me isn't hiring someone who read about shipping AI — it's hiring the person who ships and operates it alone, every day.
How we engage
Every engagement is scoped and quoted in writing after a short call — fixed scope, never open-ended hours. Common needs have a fixed price — a $497 AI Quality Audit, packaged builds from $4,997, and monthly care from $299/mo; larger builds are scoped and quoted in writing after a short call. The audit fee credits straight into the build if you continue.
Proof — built & operated
Two live products, one engineering system, one operator. Every figure below was counted from source code; you can open either front door.
Nexural
Sage Ideas Academy
Selected client & project work, each with evidence you can open:
| Engagement | Outcome | Evidence |
|---|---|---|
| QA platform (nexural-qa-os) | A CVE gate went red mid-release, blocked the ship, and was back to 13/13 green within the hour — a platform that refuses to show a fake green. | live capture |
| Playwright regression suite | 37/37 specs green in CI, 15.3s, zero flakes — a suite a release manager can trust. | public GitHub |
| RAG evaluation dashboard | An AI assistant that can't answer without a citation — 100% citation coverage enforced. | dashboard capture |
Track record
Before Sage Ideas, thirteen years making sure real software worked at scale:
- HighStrike (2021–2026) — quantitative finance & automation. Kubernetes-based test infrastructure against live trading workflows moving $100k+ daily; flaky rate cut 10% → <1%, test execution 45min → 8min. (figures self-reported from a prior full-time role)
- The Home Depot (2012–2021) — Fortune-50 software tester → automation → cloud engineer. Selenium/Python framework for systems serving 2,300+ stores; regression 4h → 75min, 300+ cases at 99.5% stability; led an AWS migration that cut infra cost ~35%. (figures self-reported from a prior full-time role)
Credentials & languages
ISTQB CT-AI ISTQB Test Automation Engineer ISTQB CTFL AWS Cloud Practitioner AWS ML Foundations TripleTen AI Automation
B.S. Computer Science, Full Sail · B.S. Finance, Kean · English & Portuguese native · Spanish professional. Client-facing work delivered in English, Spanish, or Portuguese.
Sage Ideas LLC · Figures counted from source repositories as of 2026; prior-role figures are self-reported and labeled. Revenue and enrollment figures are intentionally omitted. Live products linked above.
Company Overview · download PDF ↓ · see the proof → · ← portfolio