About me

I’m Dipam, a technology enthusiast, science nerd, and lover of food and travel.

I’m currently working as a machine learning engineer at AIcrowd Research. Primarily, I work on hosting research oriented competitions, especially focussed on reinforcement learning. AIcrowd has hosted many research competitions in machine and reinforcement learning. In fact, I joined AIcrowd after winning 3rd position in the NeurIPS 2020 Procgen competition hosted there. I’ve contributed heavily to the NeurIPS 2022 Citylearn Challenge, Multi Agent Behavior Challenge 2022, ZEW Data Purchasing Challenge, NeurIPS 2021 Deepracer Challenge, Flatland 3, and Multi Agent Behavior Challenge 2021.

Previously, I was at Intel, where I worked on fault detection of industrial parts. The final solution comprised on a combination of detection using supervised learning with CNNs, and prediction via unsupervised learning using autoencoders. Prior to that, I also worked on developing hardware accelerators for CNNs, based on SystemVerilog and OpenCL.

I love participating in competitions, although I have not found the time to do so recently. On Kaggle, I’ve participated in a bunch of competitions for Computer Vision, and won some silver medals. I also took part in NeurIPS 2020 Procgen Competition, which highly motivated to pursue reinforcement learning research in the future. I’m fascinated with the current pace of progress in AI, and yet it feels like there are more questions than ever. Deep reinforcement learning in particular, feels extremely promising to me, and one of the open questions around it how to make Deep RL agents generalize.

My undergraduate and masters studies are in Electronics and Communication Engineering, at NIT Rourkela, India. There, I worked hands on in robotics, primarily with embedded systems, vision and control. I was one of the first members of Team Tiburon where we developed an autonomous underwater vehicle from scratch. It is certainly one of the most challenging projects I had worked on. I learned immensely about not just robotics and product development, but also teamwork and leadership. For the team, participating in competitions was a strong motivator, and we won 2nd position in the Singapore AUV Competition (SAUVC) 2018, becoming the first team from India to win a prize at SAUVC.

I love to travel, especially places where I can go trekking or diving. Pictures speak for themselves.

Phulara Ridge Trek Boat trip Komodo Island

I’m a huge fan of many science fiction and fantasy series. Especially the Star wars and Tolkein universes, cliché, I know; but they’re so popular for a reason. 😉

Have a research idea you’d like collaborate on around machine learning or reinforcement learning? Or just want to nerd out with me about your favorite fictional universe, discuss technology, science, politics and philosophy? Feel free to reach out to me on Twitter or email. 😃