Integrate AI into your product — not just a demo.

I'm Sandeep, ex-CTO at ToDesktop (YC W20) and founding engineer at Blitz.gg. I help teams design and ship LLM copilots, RAG, agents, and production AI workflows — from strategy to working software in weeks.

Free intro call. Fixed scope for consultancy and implementation.

Previously at OP.GG, Vama, ToDesktop (YC W20), Blitz.gg, FusionCharts

Shipping production software since 2017

YC leadership, Electron at scale, and AI integrations in real products — the background I bring to consultancy and implementation work.

OP.GG

Individual Contributor

Jan 2026 – Mar 2026 · Remote

  • Prepared the entire production codebase for LLM-driven automation, restructuring for agent-readable context and safe automated refactors.

Vama.com

Senior Electron Developer

Apr 2024 – Oct 2025 · Remote

  • Configured and deployed scalable cross-platform desktop applications for Windows and macOS.
  • Engineered UI virtualization techniques that measurably improved rendering performance and app responsiveness.
  • Migrated the application IPC layer from WebSockets to gRPC, cutting latency and improving reliability.

ToDesktop.com

Chief Technology Officer

Jan 2023 – May 2023 · Remote · YC W20

  • Led technical onboarding and integration with the Cursor.ai team, optimizing their Codex-backed AI code editor for deployment inside ToDesktop.
  • Migrated core builder infrastructure from CircleCI to Microsoft Azure, improving deployment scalability.
  • Built internal developer tools that measurably increased engineering team efficiency; core contributor to the No-Code Builder platform.

Blitz.gg

Founding Software Engineer

Dec 2018 – Dec 2022 · Remote · Los Angeles

  • Built the Blitz.gg desktop application from the ground up, scaling infrastructure to 5M+ daily active users.
  • Invented Electron Delta Updates, cutting update bandwidth from 3PB to 1TB — saving over $50,000 in infrastructure cost.
  • Owned the Squirrel-to-NSIS auto-updater migration and the full macOS port, released solo.
  • Reverse-engineered Valorant's client network protocol to power a direct, low-latency in-app connection.
  • Shipped React overlays for League of Legends and led a full React/Redux/GraphQL rebuild of the Blitz.gg web app.

Granicus.com

Software Development Engineer II

Sep 2018 – Nov 2018 · Bangalore

  • Led the frontend team through a full rebuild of EngagementHQ using React, Redux, Thunk, and Next.js.

FusionCharts.com

Software Development Engineer

Jul 2017 – Aug 2018 · Bangalore

  • Core contributor to MuzeJS, an open-source data visualization library.
  • Built FusionBoard's backend, unifying Java and PrestoDB across heterogeneous databases and flat files.
  • Architected FusionExport's core export system and built FusionExport Desktop (Electron); authored FusionTime bindings for AngularJS, Angular 2, React, and jQuery.

From AI idea to production integration

I advise on what to build, then implement it — Cursor and Codex workflows, LLM automations in production codebases, copilots, RAG, and agents that ship to real users.

Talk through your AI integration

AI integration strategy

Map where AI creates real leverage in your product or ops — models, data, evals, and a build plan your team can execute.

LLM copilots & chat

Production copilots and assistants on OpenAI, Anthropic, or open models — streaming, memory, tool use, and guardrails.

RAG & knowledge systems

Ground answers in your docs, tickets, and internal data so teams get accurate responses instead of hallucinations.

Agents & workflow automation

Multi-step agents that take action — draft content, update CRMs, run reports, and handle ops tasks end to end.

A simple, transparent process

01

Intro call

15 min

Share where AI should fit — product feature, internal ops, or full integration. I'll say if I'm the right fit.

02

Integration plan

2–3 days

Fixed scope covering models, data, architecture, timeline, and price — consultancy and implementation in one plan.

03

Implement & ship

4–12 weeks

Weekly demos of working AI features in your stack. Eval loops, guardrails, and production readiness built in.

04

Handoff & support

2–4 weeks

Docs, runbooks, and a support window. Retainer available if you want ongoing AI product work.

Trusted by founders and engineering teams

“Sandeep is an outstanding engineer. During our time working together at Blitz, he consistently demonstrated a deep understanding of Electron and React showcasing an impressive ability to create cross-platform applications very quickly.”

Adil Virani
Adil Virani
Cofounder & President Blitz.gg, Forbes 30 under 30

“Sandeep is diligent, hard-working, honest and has great communication skills. I was super glad to have the opportunity to work with him at ToDesktop and hope to be able to work with Saneep again in the future.”

Dave Jeffery
Dave Jeffery
CEO of ToDesktop.com (YC W20)

“Sandeep is a skilled full-stack developer who prioritizes client outcomes. I highly recommend working with Sandeep!”

Jon Hainstock
Jon Hainstock
Co-founder & CEO of ZoomShift

“Sandeep was extremely professional and went above and beyond to make the project deliver on time. I really respect his skill set and strongly recommend to anyone looking for professional help with anything related electron.”

Satya Avasarala
Satya Avasarala
Co-founder DailyBee

Ready to integrate AI?

Book a free 15-minute intro call. We'll talk through your use case, data, timeline, and whether I'm the right fit for consultancy and implementation.

Reply within 24 hours · No obligation · Worldwide