I'm Madhur Goel. I build hard tech.
18 years. Two tracks. A lot of problems solved.
I'm a consultant specializing in AI/ML systems and low-latency distributed infrastructure. I've built High Frequency trading platforms, shipped generative AI to real consumers, deployed real-time video infrastructure across 6 regions, and am currently deep in the LLM and agentic AI wave. I work with enterprises and deep-tech startups who need someone who's done it before.
The story
My background is in biochemical engineering and biotechnology from IIT Delhi : not CS. Every skill I've built since 2006 has been earned by solving hard problems. That path is the consulting practice.
| Year | Role | Key thing |
|---|---|---|
| 2006 | Research Intern, UCLA | Ported MATLAB to C : performance as a feature. Days → minutes. |
| 2008 | Product Associate, iXiGO.com | Meta-search handling 20+ concurrent queries per user. First job. |
| 2010 | Quant Strategist & Tech Lead, HFT prop desk | HFT strategies, NSE market connectors : first deep low-latency work. |
| 2015 | Deep learning consultant | Video, NLP, anomaly detection; Netflix server load prediction (subcontract). |
| 2017 | Co-founder & CTO, Tangent.ai | Enterprise-scale GANs; first Fortune 500 fashion client; Alchemist SF. |
| 2020 | Co-founder & CTO, webrtc.ai | Globally deployed video confrencing for low-bandwidth markets; AI in the browser. |
| 2022 | Founder, SelfAware Machines | High Frequency Trading platform from scratch; kernel Bypass and cache hits. |
| 2025 | Fractional CTO, Nextgen Techno Ventures | Payment orchestration on server infrastructure based on dynamic variables. |
| 2025 | Consultation Work, SAM42.COM | LLM Systems & Agentic Pipelines with prompt optimization strategies |
The two tracks
AI/ML systems: From training StyleGANs on proprietary data to RAG with DSPy : full-stack AI for production.
Low-latency infrastructure: From NSE certification to 6-region AWS : systems where milliseconds matter.
Why both: The hardest problems sit at the intersection. Real-time AI needs both. Most consultants are strong in one : I operate across both.
Credentials & education
- IIT Delhi : 5-year Dual Degree (B.Tech + M.Tech), Biochemical Engineering and Biotechnology, 2006. Not a CS degree : skills built through delivery.
- Alchemist Accelerator : Cohort XX, San Francisco, 2018–2019. Funded, first enterprise client, technical moat for Tangent.ai.
- NSE certification : Algo approval for proprietary trading platform, 2022–2025.
- Research : UCLA MCBD: protein functional linkages via PFAM domain pattern analysis.
The consulting practice (sam42)
After 18 years building deep-tech products : inside startups, enterprises, and as a founder : I'm offering that experience as a consulting resource for teams with hard technical problems.
Who I work with: Enterprises with complex AI/infra problems, deep-tech startups needing senior bandwidth, investors evaluating technical risk.
What you get: Direct access to the principal, fast turnaround, shipping mindset over slide decks.
Availability
Currently available for new engagements.
I'm selective : I work with clients who have a real problem and the commitment to solve it. If that sounds like you, let's talk.
Get in touch →