graph by datavruti
mygraph.id/mayanksingh

Mayank Singh

Senior Engineering Lead / AI Platform Architect

Gurugram, Haryana 122001, IndiaBuilding since Jun 2018 (8+ yrs)
LinkedIn
Headline

I spent four years hardening core distributed systems and full-stack architecture skills before redirecting that same rigour toward LLM orchestration, code intelligence, and multi-agent systems at real production scale.

cut LLM costs by ~30-40%Cost per Request
~40,000PRs Reviewed Daily
850K+Marketplace Installs
15+LLM Providers Unified
Represented Roles

Specialized Functions & Roles.

Engineering capabilities and specialized roles demonstrated across real production work.

AI Platform Architect

Primary Role

Senior Engineering Lead

Distributed Systems Architect

Agentic AI Engineer

MLOps / Platform Engineer

Signals

Live in Production.

Each signal ties back to a specific, evidenced moment.

Built

3 moments

Systems shipped end-to-end, from schema to deploy.

  • •Building an end, to, end autonomous coding agent from Jira ticket to PR
  • •Built the multi-agent orchestration core for Wingman
  • •Sole-architected an IDE assistant with 850K installs

Scaled

2 moments

Taken from prototype to real production volume.

  • •Scaled a code knowledge graph to a 1,000-repo target
  • •Scaled an AI review pipeline to 40,000 PRs a day
Growth trajectory

Mayank's career, mapped.

A timeline of milestones captured directly from real work.

  1. Started architecting core platform systems

  2. Took on real-time and offline-first UI

  3. Owned auth architecture end-to-end

  4. Moved into AI platform architecture

  5. Became sole architect of core products

  6. Designed multi-agent orchestration core

  7. Scaled a code knowledge graph platform

Stack

Daily tools.

What Mayank actually reaches for, pulled from the tags attached to captured moments.

Work

Key Projects By Role

Case studies, not bullet points - impact numbers attached to each one.

Bito Inc.

Senior Engineering Lead / AI Platform Architect · Feb 2022 - Present
2022

IDE Plugins & Web UI

Sole architect of Bito's AI Code Assistant for VS Code and JetBrains, and of the primary SPA covering workspace management, auth, billing, and role-based administration.

850K+ marketplace installs
Stack:TypeScriptAngularVS Code Extension APIJetBrains SDK
View case study
2023

Code Review Agent (CRA) & Adaptive AI Review

Adaptive-learning review pipeline processing ~40,000 PRs/MRs per day across GitHub, GitLab, Bitbucket, and Azure DevOps, using a Go/Kafka service turning feedback into OpenAI embeddings refined via Qdrant vector search.

~40,000 PRs/MRs processed per dayUnified across 4 major Git providers
Stack:GoKafkaQdrantOpenAI embeddings
View case study
2024

Bito Wingman

Agentic AI platform unifying CLI, Web UI, and REST/WebSocket into one orchestration core across 15+ LLM providers, with multi-agent orchestration, sandboxing, and an extensible Hooks to Skills to Plugins framework.

Unified orchestration across 15+ LLM providersArchitected a multi, agent orchestration core that runs 10 concurrent worker agents by default in coordinator mode, with configurable concurrency limits (sub, agent spawning defaults to 2, auto, raised to 5 under coordinator mode) to scale with available system resources.
Stack:MCPMulti-agent orchestrationWebSocketGit worktrees
View case study
2024 - 2025

XMCP (Code Intelligence System)

Custom MCP server powering a cross-repo code knowledge graph used by AI clients like Claude and Cursor, indexing ~250 repos across 15-20 tenant orgs with multi-tenant OAuth 2.1/PKCE and SSO, and Temporal-orchestrated indexing.

~250 repos indexed across 15-20 tenant orgsDesigned to scale to 1,000+ repos
Stack:MCPNode.jsTemporalS3
View case study

Senpiper Technologies Pvt Ltd

Software Engineer (SDE I → SDE III) · Jun 2018 - Jan 2022
Outside Work

Side projects, open source & publications.

Independent builds, open-source contributions, technical writing, and experiments beyond the day job.

Side project

Autonomous Dev Cycle Agent

Building toward a development cycle that runs itself end to end, taking a Jira ticket in and producing a PR out with minimal human intervention. Currently focused on getting an agent to run autonomously for 8, 9 hours without a human in the loop, tackling state checkpointing, knowing when to stop and ask versus guess, recovering from bad decisions made hours earlier, and keeping context from degrading over long runs. Most of it is still POC stage and many attempts have failed, but the failures have been informative.

Learning

Formal record.

  • B.Tech, Computer Science & Engineering
    Galgotias University, Noida, UP
    2014 - 2018
  • XII
    CBSE, Bareilly, UP
    2012 - 2014
  • X
    CBSE, Bareilly, UP
    2010 - 2012
Activity heatmap

A record of momentum.

A combined view of Mayank's activity across the tools used daily.

LessMore

Profile last updated today.

Liked Mayank's graph?

Build yours today.

Start with your resume