Hi, I'm Kushal Agarwal

CS Master's Student @ UIUC | Software Engineer | AI/ML Researcher

About Me

Hello! I'm Kushal, a Master's student in Computer Science at the University of Illinois at Urbana-Champaign (expected May 2026). I graduated with my Bachelor's in CS from UIUC in May 2025, earning Dean's List and James Scholar honors with a 3.8/4.0 GPA.

I'm passionate about building scalable systems and leveraging AI/ML to solve real-world problems. My experience spans from biomedical AI research to building high-throughput backend systems at Amazon. I am currently working on a healthcare platform for the Indian market that is achieving ABDM compliance.

I specialize in full-stack development, cloud architecture, and AI/ML engineering, with a focus on creating efficient, secure, and user-centric solutions that make a meaningful impact.

Skills & Expertise

🤖

AI/ML Engineering

GraphRAG, PyTorch, Agentic Workflows, JSON Mode, Prompt Engineering, Raw Python Scripting

💻

Languages

Python (Advanced), TypeScript, Java, SQL, C++, Kotlin

☁️

Backend & Cloud

AWS (Lambda, DynamoDB), Supabase, PostgreSQL (RLS/RBAC), Edge Functions, Docker

🛠️

Tools & Platforms

Git, Datadog, Jira, Google Colab, RESTful APIs, CLI Tooling, Flutter

Experience

Software Development Engineer Intern

Amazon

June 2025 – August 2025
  • Reduced client onboarding time from ~1 week to zero-touch by automating end-to-end workflow using AWS Lambda, DynamoDB, and S3
  • Increased onboarding throughput by ~6x with a stateless, idempotent architecture utilizing exponential-backoff retries for resilience
  • Ensured 100% data consistency during legacy migration by writing custom Python and TypeScript utility scripts
  • Eliminated 95% of engineering support tickets by exposing internal business logic via secure RESTful APIs

Research Assistant (Bio-Medical AI)

Data Mining Group, UIUC

August 2024 – May 2025
  • Boosted model question-answering accuracy by 5% on clinical benchmarks using agentic workflows in Python
  • Achieved 4% accuracy gain by implementing GraphRAG to transform unstructured biomedical text into semantic triples/graphs
  • Optimized context window usage and reduced hallucinations with custom retrieval pipeline bypassing high-level abstractions
  • Validated system performance improvements by writing custom Python evaluation scripts for benchmarking

Featured Projects

🏥

Aria (ariamed.ai)

Digital prescription platform for the Indian market achieving ABDM compliance. Standardized drug data with SNOMED International terminology and implemented secure multi-tenant backend with RBAC.

Supabase PostgreSQL Edge Functions Flutter TypeScript
🤖

Bio-Medical AI Research

Boosted clinical question-answering accuracy by 5% using GraphRAG to transform biomedical text into semantic graphs. Optimized context window usage and reduced hallucinations with custom retrieval pipeline.

Python GraphRAG JSON schemas PyTorch
🏎️

Illini Motorsports Data Systems

Built Python data ingestion pipelines to parse and normalize Bosch DARAB sensor logs. Reduced gear-shift times by 0.05 seconds per change using algorithmic telemetry analysis.

Python Data Pipelines Time-series Analysis

Get In Touch

I'm always open to discussing new projects, creative ideas, or opportunities to be part of your vision.