01Work 02About 03Writing 04Skills 05Background 06Contact

Ayooluwa Obembe

QA Automation & AI Reliability Engineer — LLM, RAG, and legal-tech systems.

I build CI-ready test automation, LLM/RAG evaluation harnesses, and tool-calling QA pipelines for SaaS, fintech, and legal-tech products.

Lagos, Nigeria · Available for remote work.

View work → Contact
Ayooluwa Obembe
01 About

I'm a QA engineer and Barrister at Law working at the intersection of software quality, AI reliability, and legal reasoning. I build systems that don't just produce fluent output, but can be tested, audited, traced, and trusted — the properties that matter most in high-stakes, regulated work.

My central project is LexTutor, a legal-AI exam-intelligence system for Nigerian Bar Finals candidates, designed around route-validated grading, authority discipline, and reasoning fidelity rather than raw generation. It grows out of a simple conviction, argued in my paper The Reliability Gap: that AI's failure on legal reasoning is architectural, not a matter of scale.

Alongside it I build the testing infrastructure that holds AI systems to that standard — including TestAIgnite, a framework that pairs CI automation with grounded model-driven root-cause analysis.

Role
QA & AI Reliability Engineer
Open source
Creator of TestAIgnite
Domains
Fintech · RegTech · Legal-AI
02 Writing
The Reliability GapWhy Artificial Intelligence Still Cannot Be Trusted with Legal Reasoning Position Paper

A field analysis of the gap between fluent legal language and reliable legal reasoning. Argues that the failure of general-purpose AI on legal questions is architectural — a lack of navigational discipline, not a lack of intelligence — and defines the properties any credible legal-reasoning system must satisfy: authority grounding, procedural fidelity, commitment discipline, honest boundaries, and traceability.

Read paper →
03 Selected Work
TestAIgnite npm · live

Published npm package for autonomous QA intelligence: multi-model LLM orchestration, tool-calling failure analysis, and CI-ready reporting for test suites.

Node.js · LLM Tool Calling · Cypress · CI/CD View on GitHub →
LexTutor Private, Pre-Launch

AI exam-performance system for Nigerian Law School Bar Finals candidates. Designed around a curated archive of over a decade of Bar Finals questions with verified legal authorities, structured legal knowledge mapping, and route-validated grading — validating the legal path, not just the answer — to recommend targeted drills.

Legal AI · Bar Finals · Knowledge Graph · Route Validation · Exam Intelligence · Adaptive Learning
RegWatch Automation RegTech

Risk-focused automation and quality gates for RegTech compliance workflows serving Tier-1 financial institutions.

Playwright · API · CI/CD · RegTech View on GitHub →
API Detective framework

Enterprise integration-testing framework that turns API flows into repeatable CI signals across Postman, Newman, PowerShell, and Jenkins.

Postman · Newman · PowerShell · Jenkins View on GitHub →
04 Capabilities
Test Automation
Cypress · Playwright · Pytest · Appium · BDD/Gherkin
AI & Advanced QE
LLM eval harnesses · RAG evals · Tool-calling workflows · Agent workflows · Prompt regression · AI failure triage
API & Performance
Postman (Newman) · REST · K6 · Contract testing
Languages
JavaScript · TypeScript · Python · Java · PowerShell
CI/CD & Tools
GitHub Actions · Jenkins · Docker · Jira · CodeQL · CI quality gates
05 Background
Dec 2025 — Present
Lead AI Systems & QA Engineer
Opex Consulting — RegTech365
Dec 2023 — Dec 2025
Software QA Engineer — Automation & API Testing
Walure — Remote
2024 — 2025
Freelance / Contract QA Engineer
Testlio · BBC News · Matterport · FELLI Music · Wellcrafted
Education
Barrister at Law — Nigerian Law School, 2025
LL.B, Redeemer's University, 2022
06 Contact

Let's build reliable AI and QA systems.

LinkedIn → github.com/sphinx010 → Download Résumé ↓