Anubhav Sachan
Anubhav Sachan

Learning with experience replay!

Software Engineering & Machine Learning

Hey there!

I love solving *complex* software engineering problems!1

A deep (neural) pull towards machine learning and artificial intelligence — drives me to *bridge the gap* between research and production, from whiteboard sketches to scalable systems that people can actually use.2

And I still have an *affinity* for conversations about the fundamentals of reinforcement learning — the state-action space and the reward functions.

Since you're curious, my experience, research & projects sections trace the fields I've wandered through — and I occasionally write3 too — you are welcome to drop a hello!

Experience
SDE II at Samsung Research (SRI - Delhi)

Currently leading the architectural design of backend systems and core models for Automated Defect Analysis and Conversational AI. I've spent the last 5 years engineering an AR-assisted TV Installation platform2, multi-cloud Generative AI gateways and omni-channel CRMs, following research internships at NVIDIA and Saarthi.ai.

View all
Research
Patents & Publications

Designed the core technical architectures for 7 patents across Conversational AI, Generative Video, and Extended Reality (XR) — spanning innovations from context-aware virtual assistants to dynamic shadow transformations — alongside peer-reviewed Reinforcement Learning research for optimized energy harvesting presented at IEEE GLOBECOM.

View all
Projects
Applied Deep (& Reinforcement) Learning

Explored advanced architectures across NLP, RL, and Computer Vision — spanning Graph Convolutional Networks (GCN), federated learning POCs, and offline multi-agent dialogue policies, alongside unsupervised domain adaptation and published open-source tools4.

View all
Highlights
  • Stellar Project Award (Q1 '26) for Best Growth (Innovation) — for AR Assisted TV Installation System2 at SRI - Delhi.
  • Challenge Project Award ('24) — for Generative AI-based features in CRMs at SRI - Delhi.
  • 7x Innovation Award Winner (for each patent) at SRI - Delhi.
  • Young Achiever of the Year ('22) at SRI - Delhi.
  • Reinforcement Learning based Bachelor's thesis with UNSW Sydney, presented at IEEE GLOBECOM '21.
  • Research Internships at NVIDIA (Spring '21), Saarthi.ai (Summer '20), and IIT Indore (Summer '19).
  • Winner, IEDC & UGRC Grants ('20) (worth ₹1.25L).
  • National Level Finalist, Smart India Hackathon ('19)5 & Winner, NIT Conclave Hackathon ('19)6.

1 Using data structures and algorithms (sometimes!) ↩
2 Have patented my latest work in Samsung! ↩
3 On Medium, mostly about neural networks. ↩
4 Published as a PyPI package ("bingscraper"), cited on Google Scholar. ↩
5 National Level Finalist, Smart India Hackathon '19 (for ezDI's problem statement) for "Patient Case Similarity". ↩
6 Winner, Hackathon, NIT Conclave '19 (amongst all 31 NITs) for "Paperless". ↩