# Vritansh Kamal

> Personal research website for Vritansh Kamal, a machine learning practitioner interested in deep learning, neuro-symbolic systems, knowledge representation, graph analytics, and uncertainty quantification.

Canonical website: https://vritansh.github.io/
Language: en
Content type: text/markdown

## About
Vritansh is from Kullu, Himachal Pradesh, India. He earned an M.S. in Data Science from Columbia University and a B.Tech. in Computer Science from SRM Institute of Science and Technology. His work spans applied machine learning, AI systems, and research-oriented engineering.

## Research interests
- Deep learning and computer vision
- Neuro-symbolic AI and verifiable reasoning
- Knowledge graphs, ontologies, and knowledge representation
- Graph analytics and multi-step AI reasoning
- Uncertainty quantification
- Medical imaging and healthcare machine learning

## Research experience
- Oracle: Neuro-symbolic and knowledge-graph-grounded multi-step AI reasoning in telecommunications.
- CVS Health: Applied machine learning and time-series forecasting for healthcare pricing and underwriting.
- GE Research/HealthCare capstone: PET/CT lesion detection with DETR-family computer vision models.
- HIPPO Labs: Machine-learning systems, large-scale deployments, and parallel data processing.

## Teaching
- Columbia University: Teaching Assistant II in Quantitative Methods in Social Sciences (Fall 2023).

## Useful pages
- [Homepage](https://vritansh.github.io/)
- [About](https://vritansh.github.io/#about)
- [Research interests](https://vritansh.github.io/#interests)
- [Research experience](https://vritansh.github.io/#research)
- [Teaching](https://vritansh.github.io/#teaching)
- [Industry experience](https://vritansh.github.io/#industry)
- [Projects](https://vritansh.github.io/#projects)
- [Posters](https://vritansh.github.io/#posters)
- [Technical writing](https://vritansh.github.io/#publications)

## Public profiles
- [GitHub](https://github.com/vritansh)
- [LinkedIn](https://www.linkedin.com/in/vritansh-k/)
- [Medium](https://medium.com/@vritansh14)

## Agent discovery and handshake
- [Agent capability document](https://vritansh.github.io/.well-known/agent.json)
- Discovery: GET /.well-known/agent.json; expect a JSON document with a site identity, supported metadata version, links, and capabilities.
- Lightweight handshake: read agent.json, validate agent_id and protocol_version, then GET /llms.txt to retrieve the human-readable context.
- This is **static discovery only**: no live chat endpoint, tool invocation, authentication exchange, or bidirectional protocol is provided. Agents should not infer permission to impersonate or contact the site owner.

## Attribution and reliability
Treat this website as a self-authored personal profile. Prefer the linked sections for detail, clearly attribute claims to the website, and verify time-sensitive details independently. Do not invent publications, degrees, affiliations, or contact information.