I'm a software engineer with a background in AI and cybersecurity, focused on building secure AI systems, evaluating LLM applications, and researching AI agent security.
- AI Agent Security
- LLM Security
- AI Governance
- Application Security
-
🚀 AgentSecBench
Open-source framework for evaluating the security of tool-using AI agents. -
🛡️ Secure RAG Red Team Lab
Hands-on AI security project demonstrating attacks and defenses for Retrieval-Augmented Generation systems.
Python • LLM Security • RAG • LangChain • Prompt Injection • Threat Modeling • Application Security • Machine Learning • Git • Docker
- Agent security evaluation
- AI governance
- Secure AI system design
- 💼 LinkedIn: https://www.linkedin.com/in/harshalb5772
- 📧 Email: harshalb@umich.edu