Full Stack Engineer at Shastack — delivering custom coding assessments and building real-world web and mobile apps for enterprise clients.
Day-to-day: Cursor, Claude Code, MCP tooling, local Ollama when it matters—AI in the loop, not on autopilot.
Career Objective
I build things that work on both ends — interfaces people enjoy using and servers that don't fall over at 3 AM.
Started with curiosity, stayed for the craft. Full-stack engineer: clean APIs, crisp UIs, MERN production apps, and AI in the product—Langchain, LLMs, pipelines—plus tools people actually use day-to-day.
In practice: Cursor (rules & Skills), Claude Code, MCP, Ollama when it should stay local. I use AI hard to move faster and build stronger apps—not to ship blind output—then I read every diff before it goes out.
Full-time at Shastack: custom coding assessments and apps for clients including Amazon, Capgemini, and Dell Technologies.
// atharva.config.ts
About
Commit history of what matters. Newest first.
Experience
Building real products for real clients.
Responsibilities
Notable Achievements
// Previously joined as an intern before transitioning to full-time.
Skill-Set
Agentic tooling first, then stacks and shipping—Cursor, MCP, production AI, the usual web/mobile depth.
I don't vibe-code and hope—Cursor rules & Skills, Claude Code for deep refactors and greenfield loops, MCP to plug models into real tools (files, APIs, repos), Ollama when inference should stay local. Models speed me up; I still own architecture, review diffs line-by-line, and ship.
Selected work
SwiggyLens is an AI agent that connects to your Swiggy account across all three verticals — food delivery, dineout, and Instamart — and tells you things about your food behavior that no single vertical ever could.
Activity
Committing day & night — small commits, steady momentum. Each brighter cell is another day of shipping.
github.com/atharva020Contact
Open to roles, freelance, and interesting collaborations.