Skip to content
Pushkar.

Pushkar.

0%

Find me on GitHub LinkedIn

Ask me anything.

Ask Pushkar

scripted · from my résumé

Pushkar Patil is a full-stack engineer who builds fast, AI-powered products end to end — from system design to the last pixel.

Most recently he was Lead Full-Stack Developer at IEJL in Newport Beach. Before that he was a founding engineer on StudyFetch's Honen platform, an AI learning product built with NVIDIA for K–12 education.

He holds an MS in Computer Science from Indiana University, is based in Navi Mumbai, and is open to relocation.

What has he shipped?

A few highlights:

• IEJL — a production Next.js + Supabase app: Redis caching and RPCs cut responses from 300 ms to 95 ms, Lighthouse 99, LCP under 0.6 s, and a semantic job search on Pinecone that went from 310 ms to 15 ms.

• StudyFetch — shipped the Honen pilot in a month, built workers handling 1,500+ concurrent AI generation jobs, a Redis queue for course ingestion, and company-wide SSO.

• Radical Squares — led the MVP frontend; virtual scrolling and cursor pagination took network loads from 500 ms to 10 ms.

• Side projects — MediaKit Lab, a creator-economy SaaS built solo, and MarkForge, a Rust/WASM markdown editor rendering large docs in under 16 ms.

What's his stack?

• Languages — TypeScript, JavaScript, Python, C#, Java

• Frontend — React, Next.js, React Native, Redux, shadcn, WASM

• Backend — Node.js, Express, Django, ASP.NET, REST, WebSockets

• Data — PostgreSQL, MongoDB, Redis, Firebase, Pinecone

• Cloud — Vercel, AWS, GCP, Docker, CI/CD

• Architecture — microservices, message queues, RAG, SSO, rate limiting

Where has he worked?

• IEJL · Newport Beach — Lead Full-Stack Developer (2026)

• StudyFetch · Beverly Hills — Full-Stack Engineer (2025–26)

• Radical Squares · Chicago — Software Engineer Intern (2025)

• Reva Technologies · Navi Mumbai — AI Intern (2022–23)

What's he like outside work?

Gym most days, beaches whenever he can, and a real soft spot for perfumes. He also loves sketching system architecture and building products for the fun of it.

Skills.

  • Languages (Dumbbell rack): JavaScript, TypeScript, Python, C#, Java
  • Frontend (Battle ropes): React, Redux, Next.js, React Native, HTML, shadcn/ui, CSS, WASM, TensorFlow.js
  • Backend (Bench press): Node.js, Express, Django, ASP.NET, RESTful APIs, WebSockets
  • Databases (Plate storage): MongoDB, PostgreSQL, Redis, Firebase, Pinecone
  • Cloud & DevOps (Pull-up rings): Vercel, AWS, GCP, Docker, CI/CD, Git, CDN
  • Architecture (Power cage): Microservices, Micro-Frontends, MVC, OOP, Agile, Webhooks, RPC, SSO, Rate Limiting, Message Queues, JWT, Blob storage
  • AI Tools (Kettlebells): Claude Code, Cursor AI, MCP Server, Agentic Workflows, RAG

The road so far.

  1. AI Intern · Reva Technologies

    Mar 2022 – Mar 2023 · Navi Mumbai, India

    • Engineered a CNN-based genetic-algorithm model that boosted prediction accuracy by 10%, reaching 90%.
    • Developed a software defect prediction pipeline in Python: preprocessing, model training and evaluation.

  2. Software Engineer Intern · Radical Squares

    Jun – Sep 2025 · Chicago, IL

    • Built the MVP frontend single-handedly with React, TypeScript, Tailwind CSS and Redux.
    • Introduced virtual scrolling and cursor-based pagination, slashing network load times from 500 ms to 10 ms.
    • Optimized frontend performance with polling strategies, debounce patterns and Redux state management at scale.

  3. Full-Stack Engineer · StudyFetch

    Sep 2025 – Jan 2026 · Beverly Hills, CA

    • Designed and delivered the pilot platform within a month on Next.js, TypeScript and MongoDB.
    • Optimized AI generation with Vercel Workflows, the AI SDK and RAG, plus Redis KV caching that cut load times by 80%.
    • Built Express + GCP background workers handling 1,500+ concurrent AI generation jobs, reducing timeouts by 95%.
    • Architected a Redis MQ pipeline (S3 uploads, text extraction, Pinecone embeddings) and shipped company-wide SSO.

  4. Lead Full-Stack Developer · IEJL

    Apr – Jun 2026 · Newport Beach, CA

    • Supabase RPC and Redis caching cut average response time from 300 ms to 95 ms (68%).
    • Partial Prerendering, streaming SSR and smart caching for a Lighthouse score of 99 and LCP under 0.6 s.
    • CDN delivery with ISR and on-demand revalidation cut page load times by 65% and bandwidth by 72%.
    • AI job search on Pinecone (dense + sparse) with OpenAI embeddings and query caching: 310 ms down to 15 ms.