Role: Internal AI Tooling Engineer

Led the development of a secure, internal LLM assistant for a 50-person organization, bridging the gap between enterprise security needs and generative AI capabilities.

Technical Stack: Python (FastAPI, PySide6), SQL Server (SQLAlchemy), LangChain, OpenAI API

Key Engineering Achievements:

• Secure AI Architecture: Built a Windows desktop client (PySide6) coupled with an always-on FastAPI backend, enabling employees to access internal data (news, Excel, fund docs) securely.
• Governance & Audit System: Implemented a robust Role-Based Access Control (RBAC) system with versioned client distribution. Engineered comprehensive audit logging (usage, IP, MAC address) to meet financial compliance standards.
• Leakage Prevention: Developed a keyword-based risk detection system (Grade A/B/C thresholds) that triggers admin alerts to prevent sensitive data leakage.
• Prompt Engineering Library: Deployed a curated library of ~10 prompt templates with attribution and moderation features to standardize AI usage across the firm.