Open to opportunities · Richmond, VA
Aiswaryaa Velumani
Engineer building intelligent systems

I build things that are both intelligent and reliable. From RAG pipelines and agentic systems to distributed backends and real-time infrastructure — end to end, production ready.


About

A little about me

I recently completed my MS in Computer Science at UIC (GPA 3.78), where I spent the last year building AI systems I would actually want to use in production.

My background is full-stack, which means I do not just prompt models and call it a day. I think about the whole system: how data flows in, how embeddings are stored, how answers get evaluated, and how the frontend makes it feel seamless. That end-to-end view is what I bring to every project.

Originally from Chennai, India, I am now based in Richmond, Virginia, and genuinely excited about where AI and engineering intersect. Good systems should be invisible — you only notice them when they break.

6
production projects shipped
3.78
GPA, MS Computer Science
3
LLM APIs integrated
500+
papers indexed in ResearchOS
Aiswaryaa Velumani in Chicago

Projects

Things I’ve built

A mix of AI systems and full-stack engineering. Each one is something I’d put in production.

🧠

Retention

AI Spaced Repetition Engine

Turns PDFs, YouTube transcripts, and plain text into adaptive flashcards grounded in your exact material, not generic questions from the internet.

RAG pipeline with 4 to 6 second processing and 768-dim embeddings in Supabase pgvector
SM-2 spaced repetition built from scratch across 50 to 200 concepts per user
Claude API evaluates answers for conceptual understanding, not just string matching
ReactFastAPIpgvectorClaude APIGemini EmbeddingsDocker
🔮

AgentOps

AI Agent Observability Platform

Wraps any LangGraph, CrewAI, or custom agent with one line of code and gives you full visibility into what it is doing, where it is failing, and why.

React Flow execution graph with step-by-step replay of any agent run
Automated detection of 4 failure types: hallucinations, loops, retrieval errors, token overruns
LLM as judge via Claude API to catch prompt regressions across versions
React FlowFastAPIPostgreSQLLangGraphClaude API
🔭

ResearchOS

Scientific Consensus Intelligence

Instead of returning a list of papers, this tells you what the research actually agrees on, where it conflicts, and exactly why those conflicts exist.

Indexes 500 plus papers across Semantic Scholar, PubMed, and arXiv
Classifies contradictions across 4 dimensions: methodology, population, funding bias, temporal shift
Neo4j Aura knowledge graph for citation and contradiction relationships
pgvectorNeo4j AuraClaude APIPubMedarXiv

TaskFlow

Distributed Job Queue and Worker Dashboard

A distributed job processing system built from scratch handling 1,000 plus jobs per minute with zero message loss, retry logic, and a live monitoring dashboard.

Redis message queue distributing work across async workers with horizontal scaling
Real-time React dashboard with WebSocket updates for queue depth, throughput, and failures
Mirrors production patterns from Celery, BullMQ, and AWS SQS
FastAPIRedisPostgreSQLReactWebSockets

LiveBoard

Real-Time Collaborative Whiteboard

A multiplayer whiteboard supporting 50 plus concurrent users with sub-100ms latency. CRDT-based conflict resolution so simultaneous edits never overwrite each other.

Redis pub/sub WebSocket server with CRDT conflict resolution
HTML5 Canvas rendering: freehand, shapes, sticky notes, drag to move
OAuth 2.0 auth with role-based room permissions and live cursor sync
FastAPIWebSocketsCRDTReactCanvas API
🔒

AuthVault

Production Auth and Authorization SDK

A complete auth platform with JWT rotation, OAuth 2.0, RBAC, and MFA via TOTP packaged as a reusable SDK so any FastAPI app has production-grade auth in under an hour.

Redis-backed rate limiting, bcrypt hashing, and immutable audit logs
FastAPI SDK: any app integrates in 3 lines of code
React component library for login, MFA flows, and protected routes
FastAPIJWTOAuth 2.0RedisTOTP

Skills

Technologies and Expertise

The core tools and frameworks I build with every day. Hover to pause.

🤖
Large Language Models
Core AI
🔍
RAG Systems
Core AI
🪝
Agentic Workflows
Core AI
🗃
Vector Databases
Infrastructure
🔥
PyTorch
Frameworks
🔗
LangChain
Frameworks
🧬
Embeddings
Core AI
🧠
Prompt Engineering
Core AI
🤖
Large Language Models
Core AI
🔍
RAG Systems
Core AI
🪝
Agentic Workflows
Core AI
🗃
Vector Databases
Infrastructure
🔥
PyTorch
Frameworks
🔗
LangChain
Frameworks
🧬
Embeddings
Core AI
🧠
Prompt Engineering
Core AI
🐍
Python
Languages
React
Frontend
FastAPI
Backend
🐘
PostgreSQL
Databases
📦
Redis
Infrastructure
🐳
Docker
Infrastructure
🌐
WebSockets
Backend
AWS
Infrastructure
🐍
Python
Languages
React
Frontend
FastAPI
Backend
🐘
PostgreSQL
Databases
📦
Redis
Infrastructure
🐳
Docker
Infrastructure
🌐
WebSockets
Backend
AWS
Infrastructure

Experience

Where I’ve worked

Software Engineering Intern
Rane Data Centre · Chennai, India
March 2021

Education

Academic background

Master of Science in Computer Science
University of Illinois at Chicago
GPA 3.78 / 4.0Graduated May 2026
Bachelor of Technology in Information Technology
Meenakshi Sundararajan Engineering College, Anna University
Graduated Apr 2023

Contact

Get In Touch

Open to full-time roles and interesting collaborations. I’d love to hear from you.

Send a Message