About Dipankar Sarkar
Dipankar Sarkar: hands-on AI engineer and technical leader with 18+ years building production systems. AI enablement, contract AI engineering and AI delivery.
I’m Dipankar Sarkar, a hands-on AI engineer and technical leader with 18+ years building production systems. On this site I offer three kinds of engagement: AI enablement for teams that need to change how they work, contract AI engineering for teams that need senior hands, and scoped AI delivery for a defined outcome.
What I Do
Across 18+ years, I have helped organisations:
- Transform legacy systems into scalable, modern architectures — usually a phased migration off a monolith rather than a risky rewrite, so the business keeps shipping while the foundation changes underneath it.
- Optimize digital infrastructure for performance and reliability, starting with the metric that is actually costing the business money (page load, query latency, cloud spend) rather than a generic audit.
- Implement AI, ML, and data science in production, not in a notebook: model selection, the pipeline that feeds it, and the observability that tells you when it drifts.
- Develop platforms and recommendation systems that need to hold up under real traffic, not just a demo — widget platforms, personalization engines, and the infrastructure underneath both.
- Guide technical teams through digital transformation, including the parts that are more organizational than technical: hiring, code review standards, and how decisions get made once I am no longer in the room.
How I Work
Engagements start with a working session, not a slide deck: I want to see the codebase, the dashboards, and the incident history before I propose anything. From there the shape depends on what the team needs — some engagements are architecture review and a written roadmap; others are hands-on, pairing with engineers to ship the first version of a system myself so the pattern is established correctly before the team scales it. I favor small, reversible changes over big-bang rewrites, and I write down the reasoning behind a decision, not just the decision, so the team can make the next one without me.
Recent Impact
- Migrated a major media company’s job portal and entertainment platform to microservices
- Built innovative widget platforms accelerating frontend development for major property portals
- Developed AI-powered recommendation systems for e-commerce platforms
- Led product engineering transformations for P2P marketplaces
Recent Roles
- Founder, Principal — Neul Labs (2024 – Present) — High-performance AI agent infrastructure; Rust-accelerated drop-in components for production AI stacks.
- Principal AI Architect — UK fintech (Edinburgh) (May 2025 – Mar 2026) — Built fast components of the AI agent stack for financial services.
- AI Engineering Lead — Manufactured (Dec 2023 – Nov 2024) — AI agents for global supply chain optimization.
- Head of AI & Platform — Orangewood Labs (Oct 2022 – Sep 2023) — RoboGPT, AutoInspect, EdgeML platform for industrial robotics.
- Founder — Cryptuon Research (Dec 2021 – Sep 2022) — Multi-chain APIs and MPC wallets; secured $2.5M+ in pre-launch commitments.
My Approach
I believe in building for the present while planning for the future. Every engagement combines:
- Deep technical expertise
- Strategic thinking
- Practical implementation
- Knowledge transfer
- Sustainable solutions
Books
- AI for Everyday Automation (Packt Publishing, July 2026) — seven AI workflows for automating recurring work: email, meetings, research, reports, task tracking, spreadsheets, and professional writing. Written for non-technical professionals; no coding required.
- Nginx Web Server Implementation Cookbook (Packt Publishing, 2011).
Other places I write
This site covers AI enablement, contract AI engineering and AI delivery (plus fractional CTO work). The same author maintains four other sites for distinct facets:
- dipankar.name — technology leadership mandates and leadership evidence.
- dipankar.org — board advisory, executive workshops, institutional programmes and speaking.
- dipankar.cc — applied AI research and technical validation.
- desinerd.com — personal blog since 2007 (AI, DevOps, self-hosting).
Let’s Connect
See the evidence behind this work, or send a short brief describing what you need.
“Working on some of the best technologies in the world.”