Ahitagni Das

I'm an undergrad at Rice University, where I study EECS. I currently work as a researcher at the Computational Imaging Group.

I am interested in AI, mainly computer vision, startups, and tech. I am also a part of the founding team of a Seath Biotech Startup building an AI Health Copilot.

adas@rice.edu / Resumé / GitHub / Scholar / LinkedIn

Research

I am interested in Deep Learning Theory, Computer Vision, & Spatial Intelligence. I am currently working with Control-nets and Diffusion Models, with Dr. Vivek Boominathan and Dr. Ashok Veeraraghavan.

Previously, I worked as a visiting research scentist at the MIT Media Lab, where I studied colloidal robots under Dr. Deblina Sarkar. I have experience on battery materials research, at the Ajayan Group, advised by Dr. Pulickel Ajayan. I also interned at the Nanofludics Lab at IIT Guwahati, under Dr. Kalyan Raidongia, where I fabricated blue energy devices from plastic waste.

Publications

An outlook on sodium-ion battery technology toward practical application
Mingrui Xu, Xi Chen, Sreehari K. Saju, Ahitagni Das, Atin Pramanik & Pulickel M. Ajayan
The Journal of Material Sciences, 2024

Comprehensive review analyzes sodium-ion batteries as cost-effective, sustainable alternatives to lithium-ion batteries.

Ongoing Projects

BabyGPT: A Frugal Implementation of ChatGPT-2 with Triton Kernels
[Code]

Implementation of GPT-2 with transformers, causal self-attention, & GELU activation. Future extension with Triton kernels for optimized attention computation.

Klix: A Dual-AI Implementation for Automated Business Calls with Parallel ML Engines
[Code] [Website: Coming Soon!]

Klix is a dual-AI call automation platform that manages phone calls via conversational agents, providing real-time transcription, quality evaluation, and insights—eliminating time-consuming manual tasks.

Selected Projects

	Korekau: An Autonomous Rocket Recovery System Aarush Adabala, Beck Edwards & Ahitagni Das Rice Engineering and Design Showcase, 2024 & Spaceport America Cup 2024 Autonomous descent system for rocket payloads employing GPS/IMU navigation with actuator-controlled parachute steering for targeted landing from 3,000ft, featuring fail-safe spiral protocol.
	Hyperspectral Imaging Payload for Crop Analysis Ahitagni Das, Ian Schechter Spaceport America Cup 2025. 3U hyperspectral camera system using 9-axis sensor and barometer to trigger automated agricultural imaging based on altitude & orientation, with onboard neural network for crop analysis.
	TremorSense: Victim-Seeking Sensor for Earthquake Rescue Operations Jose Gnorriega, Ahitagni Das, Albert Park, Amelia Huerta, Nico Leibert Rice Engineering and Design Showcase, 2024 Portable GPR-based victim detection system using machine learning to identify survivors under 4m of rubble with 72-hour battery life and simplified binary interface for untrained rescuers.
	Electrical Energy Generation using Contrasting Interfacial Actvties of B/N-Gr Extracted from Plastic Waste Ahitagni Das, Dr. Kalyan Raidongia International Science and Engineering Fair, 2023 Devised a process for synthesizing B/N-doped graphene from plastic waste, yielding 120 mV and 0.8 µA per 6g input. First Special Award at ISEF 2023 and $60K+ scholarship to King Fahad University.

Academic Leadership

I am involved as a Design Mentor in ENGI 120: Introduction to Engineering Design, a course where students solve problems for OEDK clients through the engineering design process. Here is a video about ENGI 120. I mentored a team to develop an infant-suction monitor for pediatric patients with feeding disorders, for UT Health. I am also a CA for COMP 140: Computational Thinking using Python.

ENGI 120: Introduction to Engineering Design.

COMP 140: Introduction to Computational Thinking using Python.