name: Karamjeet Gulati
location: San Francisco, CA
education: M.S. Computer Science @ UC Davis (3.8 GPA)
focus: AI Agents, Multi-Agent Orchestration, Distributed Systems
status: Open to Software Engineer roles
Nov 2025 — Present | San Francisco, CA
Building AI-powered testing infrastructure that ships faster with fewer bugs.
- Built fullstack AI testing platform (React, TypeScript, Node.js, PostgreSQL) — 72% reduction in manual QA hours
- Engineered Playwright test automation framework with self-healing capabilities — 1,200+ test cases/deployment, 58% less maintenance
- Designed RESTful APIs and CI/CD pipelines for test orchestration across multi-tenant environments
Jan 2024 — Apr 2025 | San Francisco, CA
First engineer. Built the autonomous browser framework from zero to production.
- Architected autonomous browser agent framework (TypeScript, Next.js) — natural language to browser actions with multi-agent orchestration
- Shipped Chrome extension + MCP framework — 79.2% reliability, 400+ monthly users, 65% less manual scripting
- Launched distributed GPU inference on Kubernetes k3s — 5x throughput on SDXL, 42% latency reduction (NVIDIA 4070Ti Super)
-1e6b52?style=flat-square){kind=icon}
Feb 2022 — Jul 2023 | India
Backend systems and data infrastructure for banking at scale.
- Developed 3 internal banking web apps with MVC architecture (React, Node.js, AWS) — 30% faster dev cycles
- Optimized 54 data migration templates across Oracle & SQL Server — 87% faster extraction
- Led database optimization (schema design, stored procedures, indexing) — 45% query performance improvement
|
M.S. in Computer Science |
B.Tech in Computer Science |
Languages
Frontend & Backend
AI / ML
Infrastructure & DevOps
| Metric | Result |
|---|---|
| Manual QA hours eliminated | 72% |
| Test cases automated per deployment | 1,200+ |
| Multi-agent orchestration reliability | 79.2% |
| Monthly active users (Chrome extension) | 400+ |
| GPU inference throughput increase | 5x |
| Inference latency reduction | 42% |
| Data extraction time reduction | 87% |
| Query performance improvement | 45% |