Research
I am researching how to make AI agents more efficient and how to make them automate human work. My primary focus is autonomous agent researcher systems and autonomous coding.
Autonomous Agents for Work Automation
Efficiency, reliability, and end-to-end execution
Researching practical agent architectures that can reason, run tasks, and deliver measurable outcomes.
01.Research Interests
Autonomous Agent Researcher
Designing agents that can generate hypotheses, run experiments, evaluate results, and produce research outputs with minimal human intervention.
Autonomous Coding
Building coding agents that can plan implementation steps, write code, execute validations, and iterate toward production-ready solutions.
Agent Efficiency
Improving latency, tool usage, memory strategies, and decision quality so agents complete tasks faster with higher reliability.
Work Automation Systems
Applying agentic systems to automate high-friction workflows across engineering, operations, and customer-facing processes.
02.Applied Research Projects
Help AI
Open-source customer service agent
Help AI is an open-source Python project that uses OpenAI GPT models with LangChain to create a customer-service AI agent. It vectorizes and stores help-center data in Pinecone, then performs similarity retrieval on user queries to provide grounded, accurate responses.
Autonomous AI Agent Researcher
End-to-end autonomous research assistant
An end-to-end AI research assistant designed to take a problem statement, propose hypotheses, run structured experiments, and produce research outputs. The goal is to automate high-value research workflows with stronger consistency and speed.
Next Direction
I am continuing to push agent efficiency and autonomous execution quality in two directions:
- →Autonomous researcher agents with stronger experiment loops and result validation
- →Autonomous coding agents that can plan, implement, test, and iterate reliably