2026

AfterQuery · Pluto Experts · Terminal-Bench / Harbor

AI data, evaluation, and software-engineering benchmark environments

Evaluated model-generated software and responses across frontend behavior, APIs, databases, authentication, configuration, tests, and runtime environments.
Reproduced failures, traced root causes across code and environment boundaries, classified failure modes, and verified proposed fixes with regression checks.
Designed realistic software-engineering benchmark tasks with reproducible environments, explicit expected behavior, automated verification, and objective pass/fail criteria.
Worked on the broken-api-deploy task and upgraded it to Harbor format for practical agent evaluation.

Full-Stack EngineeringAI EvaluationDebuggingBenchmark DevelopmentAutomated Testing

2024 - 2025

Kerob Solutions (Contract)

Remote

Built a centralized internal dashboard with a Bun and Hono backend and a React, Next.js, and Tailwind frontend.
Implemented workflows for support operations, auditing records, approving requests, and monitoring system activity.
Added automated testing, CI/CD, and cloud deployment workflows, including regression improvements supported by the existing test suite.

BunHonoReactNext.jsTailwind CSSCI/CD

2025 - Present

Independent Developer / Adulis Inc.

Remote

Built an opportunity-monitoring and publishing platform that monitors sources, parses posts, extracts structured data, detects duplicates, validates content, and publishes useful opportunities.
Designed event-driven workers, APIs, PostgreSQL state tracking, and source and content administration for a system that handles opportunity expiry and processing state.

Event-Driven AutomationTelegram APIsPostgreSQLBackend WorkersAdmin Workflows