I build AI-powered systems from problem definition to working implementation — combining product thinking, software architecture, data/API integration, and AI-assisted development.

• AI Agents & LLM Applications
• Process Automation
• AI-assisted Software Development
• Data & API Integration
• Python
• Applied AI Prototyping

Discover → observe → filter → investigate emerging Solana tokens in real time.

New Solana tokens appear faster than a monitor can follow forever. The project turns that stream into a smaller, explainable working set: multi-source discovery feeds continuous observation, meaningful source-version changes are persisted in PostgreSQL, deterministic lifecycle rules reduce low-value tracking, and a read-only Observatory exposes the surviving state for bounded AI-assisted research.

What it demonstrates: multi-source ingestion · version-aware persistence · deterministic lifecycle automation · interactive data exploration · runtime telemetry · LLM tool calling and evidence-bound research · backend and frontend contract testing

→ Explore the project, architecture, live flow, and Analyst

Problem → Requirements → Architecture → AI-assisted Implementation → Testing → Evaluation → Iteration

AI coding tools are part of the implementation loop; requirements, architecture, acceptance criteria, review, and validation remain explicit project responsibilities.

Python · LLM APIs · Tool Calling · REST APIs · PostgreSQL · Git · GitHub

This profile is being curated project by project. The reusable working standard, work packages, review records, and current build status live in portfolio/.