About
Information Science undergraduate building production-shaped Applied AI systems across document intelligence and multi-user RAG. Hands-on with FastAPI, PostgreSQL/pgvector, Redis/RQ, OCR, structured LLM extraction, retrieval evaluation, security controls, and containerized backend services.
Education
-
Information Science and EngineeringInformation Science and Engineering · Oct 2024 – Jul 2028
Skills
Tools / apps / platforms
Projects
-
Secure Multi-User RAG Document AssistantFastAPI, Redis/RQ, PostgreSQL/pgvector, Groq, Docker Compose, Langfuse
Multi-user RAG assistant with authentication, uploads, retrieval, and SSE streaming. Includes asynchronous document ingestion, tenant-style isolation, secret handling, retrieval evaluation, and Docker Compose deployment.
-
LedgerLens - AI-Assisted Supplier Invoice ReconciliationFastAPI, PostgreSQL, Pydantic, Tesseract OCR, Groq, QLoRA, Qwen
Invoice-first exception-resolution service for validating document signatures, deduplicating files with SHA-256, and handling OCR/model inference through durable queued jobs. Built evidence-backed extraction and deterministic reconciliation across invoices, purchase orders, delivery receipts, and contracts with reviewer-gated decisions and audit trails.
🏆 Achievements & awards
-
Solved 330+ algorithmic problems
Solved 141+ problems on Codeforces, 60+ on CodeChef, and 100 on LeetCode.