Erons Gberaese

AI Engineer · Full-Stack

I build production software—and the systems that make AI output reliable.

I work across intelligent applications, LLM evaluation, full-stack product engineering, and automated verification—turning ambiguous requirements into software that can be tested, trusted, and shipped.

Explore selected work Download CV
5+Years building
production systems
10+Enterprise platforms
evaluated
40%Less manual effort
through AI-assisted work
4Product layers
web · mobile · API · QA
01

About

I’m an AI-focused Full-Stack Engineer with a foundation in building web, mobile, backend, and quality-assurance systems.

At Turing, I develop and evaluate applications used in AI training workflows. My work covers AI-generated code review, prompt and rubric design, adversarial red-teaming, reproducible test environments, and automated verification for complex enterprise tasks.

Beyond AI evaluation, I have shipped products across fraud detection, commerce, health technology, real-time communication, microservices, and operational tooling. I’m most effective where product thinking and engineering depth need to meet.

Think in systems Make quality observable Ship with context
02

Core expertise

From model behavior
to production behavior.

One engineering practice across the full delivery loop.

A / 01

AI engineering

Evaluation systems that turn subjective model output into evidence-backed decisions.

  • LLM evaluation and quality assurance
  • Prompt and scoring-rubric design
  • Agentic workflows, tool use and function calling
  • Adversarial evaluation and LLM red-teaming
  • AI-generated code review
  • Human-in-the-loop verification

B / 02

Product engineering

Resilient interfaces and services designed around real product constraints.

  • React, Next.js and TypeScript
  • Node.js, NestJS and Python
  • REST APIs and authentication
  • PostgreSQL and MongoDB
  • Microservices, Redis and Kafka

C / 03

Verification systems

Automation that makes complex workflows repeatable, inspectable, and trustworthy.

  • Playwright end-to-end testing
  • API and integration testing
  • Docker evaluation environments
  • Automated evidence generation
  • CI/CD and regression workflows
03

Selected work

Complex systems,
made legible.

Sanitized case studies from private commercial engagements.

These case studies intentionally omit proprietary source code, client data, and internal interfaces. They focus on my contribution, engineering decisions, and measurable outcomes.

01 / AI EVALUATION PRIVATE COMMERCIAL WORK

TURING · 2026—PRESENT

AI training and evaluation applications

Full-stack applications and automated verification systems for evaluating AI performance across CRM, ITSM, ticketing, computer-vision, and project-management workflows.

Next.jsTypeScriptPlaywrightPythonDocker

Selected contribution

  • Built reusable Next.js application architecture for evaluation environments.
  • Designed action capture, verification, and replay for observable test outcomes.
  • Created Playwright suites and automated PDF/video evidence for test runs.
  • Designed evaluation prompts, scoring rubrics, and Docker-based checks.
  • Reviewed AI-generated TypeScript, JavaScript, and Python implementations.
  • Used Codex and Cursor to build adversarial evaluation scenarios that exposed failures in reasoning, instruction following, tool use, and agentic function calling.
REPORTED OUTCOME ≈40%

reduction in manual development and correction effort

02 / FINTECH & SECURITY PRIVATE CODEBASE

FRAUDSPECT · 2025

Fraud monitoring and analyst platform

A secure platform for monitoring transactions, investigating anomalies, and managing fraud data through analyst-focused workflows.

ReactTypeScriptNestJSPostgreSQLRBAC

Selected contribution

  • Developed responsive dashboards for transaction monitoring and anomaly review.
  • Built REST APIs with NestJS and PostgreSQL.
  • Implemented role-based access control and granular permissions.
  • Connected frontend and backend systems into complete fraud-analysis workflows.
03 / COMMERCE PRIVATE CODEBASE

SPORTY GALAXY · 2024—PRESENT

Commerce and administration platform

A full-stack commerce experience with customer journeys and internal administration workflows for products, accounts, orders, and store operations.

Next.jsNode.jsPrismaMongoDBReact Query

Selected contribution

  • Built responsive, reusable customer-facing commerce experiences.
  • Implemented authentication, product discovery, cart, and account workflows.
  • Developed administration tools for day-to-day store operations.
  • Integrated server state and client state across the application.
REPORTED OUTCOME 35%

increase in user retention

04

Experience

2026—NOW

Turing

Senior Frontend Developer, QA Automation & AI Engineer

AI training applications, LLM evaluation and red-teaming, automated verification, prompt and rubric design, agentic function-calling evaluation, and AI-generated code review.

2024—2026

Estility

Full-Stack Developer

Web, React Native, Node.js microservices, commerce, internal operations, distributed systems, and engineering leadership.

2025

FraudSpect

Full-Stack Engineer · Contract

Fraud-monitoring dashboards, secure APIs, role-based access control, PostgreSQL, and full-stack TypeScript development.

2022—2024

Akoma Health

Full-Stack Developer · Contract

WebRTC teletherapy, timezone-aware booking, WhatsApp communication, HMO integrations, and backend services.

Let’s connect

Building something
intelligent?

I’m interested in AI Engineering and Full-Stack opportunities involving intelligent applications, LLM evaluation, agentic systems, developer tooling, and production software.

erons.a.gberaese@gmail.com +234 702 506 9315

© Erons Gberaese

GitHub ↗ LinkedIn ↗ Back to top ↑