VK

Neuro-Symbolic AI · Medical AI · Machine Learning

Vritansh Kamal

GitHub LinkedIn Medium
01

About Me

I am from Kullu, Himachal Pradesh, a small Himalayan valley in northern India surrounded by mountains, rivers, orchards, and a culture I remain closely connected to. It is the place I call home and a big part of how I think about curiosity, simplicity, and staying grounded.

See Kullu on Google Images

I spent several years working in industry across technology, analytics, and machine learning before moving to the United States for my M.S. in Data Science at Columbia University. Before Columbia, I completed my B.Tech. in Computer Science at SRM Institute of Science and Technology.

Outside work, I enjoy reading, learning about neuroscience, and following questions about how humans learn, reason, remember, and make decisions.

02

Research Interests

My research interests center on Neuro-Symbolic AI and Medical AI for Healthcare. I am interested in building reliable and explainable language and vision models that combine Knowledge Graphs, structured reasoning, and Deep Learning to ground multi-step decisions in verifiable evidence. In healthcare, I am especially interested in multimodal and longitudinal reasoning across medical imaging, clinical text, and structured patient data, with an emphasis on trustworthy decision support. I am also interested in reinforcement learning and efficient post-training for improving reasoning in smaller models, alongside Uncertainty Quantification, interpretability, robustness, bias, and fairness.

03

Research Experience

My primary research experience comes from industry, where I have worked on applied machine learning, language models, neuro-symbolic and knowledge-grounded reasoning, and AI systems. My healthcare and academic research experience complements that work through longitudinal healthcare modeling, medical imaging, and applied deep learning.

Primary · Industry Research Experience

Oracle

Working on neuro-symbolic and knowledge-graph-grounded approaches for reliable multi-step AI reasoning in telecommunications.

CVS Health

Worked on applied machine learning and time-series forecasting for healthcare pricing and underwriting problems.

Secondary · Academic Research

GE Research

Master's capstone on PET/CT lesion detection using DETR and Conditional-DETR for medical imaging.

HIPPO Labs

Research Assistant — worked on machine-learning systems for large-scale deployments and parallel data processing.

04

Teaching

Columbia University

Teaching Assistant II, Quantitative Methods in Social Sciences — Fall 2023.

05

Industry Experience

06

Projects

Data Discovery Pro

Built a Python library for automated exploratory data analysis, preprocessing, and dataset visualization.

Pose Estimation

Developed deep-learning pose-estimation experiments using custom CNNs and ResNet-based transfer learning.

07

Readings

Reading is one of the things I genuinely enjoy outside work. These are a few books that have stayed with me, and I would highly recommend them to anyone looking for thoughtful perspectives on people, habits, ambition, money, and decision-making.

08

Posters

Data Science Institute, Columbia University

Poster on Capstone — December 2023.

09

Publications

Medium