Skip to content
View manojmallick's full-sized avatar
🎯
Building SigMap β€” repo intelligence for AI coding agents.
🎯
Building SigMap β€” repo intelligence for AI coding agents.

Sponsoring

@analysis-tools-dev

Block or report manojmallick

Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
manojmallick/README.md
Manoj Mallick β€” Architecting the Future with AI & Code

Typing SVG

About  β€’  πŸ† Wins  β€’  Vision  β€’  Experience  β€’  Stack  β€’  Projects  β€’  Stats  β€’  Connect

LinkedIn X (Twitter) Resume Email Profile Views

πŸ† Hackathon Victories

Hackathons Won Total Prizes Competitors

πŸ… Hackathon Scale Project Award
πŸ₯‰ ElevenLabs Hack #7: v0  Β·  May 2026 58 ElevenGov   GitHub   X 3rd Place   Prize
πŸ₯‡ Airia AI Agents Hackathon  Β·  Feb – Mar 2026 1497 Guardian β€” AI-Governed Incident Response   GitHub 1st Place   Prize
πŸ₯‡ AI Agents Reloaded: Live & Labs Benelux  Β·  Mar 2026 Β· Google Γ— B. Amsterdam 100 teams Sentinel AI β€” Multi-Agent Fraud Detection Β· Google ADK + Gemini Β· 5-agent pipeline   GitHub   LinkedIn 1st Place
πŸ₯ˆ DigitalOcean Gradientβ„’ AI Hackathon  Β·  Jan – Mar 2026 2629 GradientGuard β€” DORA Compliance Intelligence Platform   GitHub 2nd Place   Prize
πŸ† Hack for Humanity | 2026  Β·  Feb 2026 775 Amsterdam Pollution Twin Copilot   GitHub Top 20 Winner

divider

πŸ‘¨β€πŸ’» About Me

I am a Professional Full Stack Developer with over 15 years of hands-on experience in software engineering, system design, and artificial intelligence. I’ve worked across frontend, backend, DevOps, and AI, building enterprise-grade systems for real-world applications β€” from large-scale microservices to cognitive AI platforms.

I’m passionate about teaching the real, modern way of coding β€” combining deep technical foundations with today’s most advanced technologies: Generative AI, Agentic systems, RAG architectures, cloud automation, and intelligent DevOps.

Whether you are a beginner exploring your first β€œHello World,” a professional improving your stack, or a researcher/developer exploring AI systems, I guide you step-by-step β€” conceptually, practically, and strategically.

🎯 Vision & Mission

Building AI systems that are not just powerful β€” but trustworthy, auditable, and production-grade.

My focus for 2026 and beyond sits at the intersection of Agentic AI, financial compliance automation, and open-source developer tooling:

πŸ€– Agentic AI in Production

Shipping multi-agent pipelines that solve real enterprise problems with production contracts, not just demos. Every pipeline I build has deterministic pre-AI logic, human-in-the-loop controls, a full governance audit trail, and zero credentials in code. Proven across incident response (Guardian), fraud detection (Sentinel AI), and compliance monitoring (GradientGuard). Next: expanding Guardian into a multi-jurisdiction DORA/HIPAA/MAS TRM compliance platform and open-sourcing the agent architecture as a reusable framework.

🏦 FinTech Compliance Intelligence

Turning regulatory obligations β€” DORA Article 11, NIS2, SOX Section 404 β€” from a cost centre into a competitive advantage. GradientGuard reduced compliance costs from €120K/year to €2,160/year for a 100-person fintech. The roadmap: live DO monitoring webhooks, multi-tenant SaaS with CISO/auditor/engineer role isolation, and pre-filled DNB and EBA regulatory reporting templates. Target market: 50–500-person EU fintechs paying Big Four firms €80–200K/year for DORA readiness.

⚑ Open-Source AI Developer Tooling β€” SigMap

SigMap is my open-source AI context engine: it scans your codebase and extracts only function and class signatures β€” no bodies, no imports β€” then writes a compact file that Copilot, Claude, Cursor, and Windsurf read automatically. Result: 97% fewer tokens, same codebase understanding. Currently at v3.3.0 with 21 releases, 21 language extractors, a VS Code extension, a JetBrains plugin, an 8-tool MCP server, and zero npm dependencies. The roadmap includes a cloud-hosted context API, team-shared context snapshots, and deeper ADK/LangGraph agent integration.

🌍 Civic & Environmental AI

Applying verifiable, cryptographically-signed AI to public-sector challenges. Amsterdam Pollution Twin Copilot proved that civic AI can be auditable by design β€” every pollution prediction is deterministically scored, Ed25519-signed, and drift-monitored. Next steps: real-time sensor streaming integration, EU Open Data compliance packaging, and a multi-city expansion framework across Europe.

πŸ“š Democratising AI Engineering

Mentoring developers and evangelising responsible AI engineering practices β€” explainability-first, version-locked, observable, and compliant by design. Teaching the modern stack: Agentic systems, RAG architectures, MCP servers, and cloud-native AI β€” through open-source, community contributions, and hands-on hackathon leadership.

divider

πŸ’Ό Professional Experience

Role Company Duration
Gen AI Specialist & Senior Full Stack Engineer ING Nov 2022 - Present
Gen AI Specialist & Senior Technical Lead HCL Nov 2020 - Present
Senior Full Stack Engineer LTI Oct 2019 - Nov 2020
Senior Software Development Engineer ABN AMRO Jan 2017 - Sep 2019
Technical Lead / Senior Developer USAA Jul 2014 - May 2017
Software Developer TCS Oct 2011 - Oct 2019

divider

πŸš€ The Arsenal (Tech Stack)

🧩 Front-End

Front-end stack

βš™οΈ Back-End & Enterprise

Back-end stack

☁️ Cloud & DevOps

Cloud & DevOps stack

πŸ€– AI, ML & Generative AI

ML stack
OpenAI LangChain Hugging Face

πŸ”¬ Data & Databases

Data stack

divider

πŸ› οΈ Featured Projects

divider

οΏ½οΏ½οΏ½ GitHub Stats

Manoj's Stats Top Languages
Snake animation eating the contribution graph
Activity Graph

divider

🀝 Connect & Collaborate

I am always open to discussing new projects, creative ideas, or opportunities to be part of your visions.

LinkedIn X GitHub Email

divider

Made with ❀️ & AI by Manoj Mallick

Pinned Loading

  1. sigmap sigmap Public

    97% token reduction for AI coding sessions β€” zero deps, 31 languages, MCP server

    JavaScript 531 37

  2. sigmap-vscode sigmap-vscode Public

    SigMap AI context engine β€” VS Code extension. 96.9% token reduction, 80.0% retrieval hit@5, 29 languages. Works with Copilot, Claude, Cursor, Windsurf, Gemini.

    JavaScript 2 1

  3. sigmap.nvim sigmap.nvim Public

    SigMap AI context engine β€” Neovim plugin. :SigMap, :SigMapQuery, statusline widget, auto-run on save. 96.9% token reduction, 80.0% hit@5.

    Lua 2

  4. sigmap-benchmark-suite sigmap-benchmark-suite Public

    Reproducible benchmarks for SigMap β€” 405 repos, 51 coding tasks, Devin agent A/B, BM25 re-ranker. ~99% fewer tokens, 96Γ— cheaper, agents ~61% faster.

    Shell 2

  5. copilot-ci-doctor copilot-ci-doctor Public

    JavaScript 2 1

  6. guardian guardian Public

    AI-Governed Incident Response for Regulated Financial Systems

    JavaScript 2 2