Hi! I’m Rushir
Bhavsar.

+ Product, AI, DevOps

+ Ahmedabad, India

+Open to Product Manager, AI Engineer & Forward-Deployed Engineer roles.

Currently building sushrutalgs.ai , a medical AI platform for India’s students, residents, and surgeons. Previously at ASU and Cadence. I also wrote Zero to Hired, a playbook for the AI job search.Looking for Product Manager, AI Engineer, and Forward-Deployed Engineer roles.

2025 – Present
Founding Engineer @sushrutalgs.ai @sushrutalgs.ai
Medical AI for India's surgical trainees: RAG backend, edge gateway, web and iOS clients.
Medical AI for India's surgical trainees: a RAG backend over surgical textbooks, an edge gateway for auth and quota, and web and iOS clients that stream cited answers. Live in production.
- Next.js 16
- Swift 6
- FastAPI
- Qdrant
- Neo4j
- Cloudflare Workers
Jul 2025 – Oct 2025
ML Engineer Intern @Cadence @OpenEye, Cadence Design Systems
Built an ESM2 + contrastive-learning pipeline; scaled protein property prediction from 150 to 50K+ sequences.
Engineered an ESM2 + contrastive-learning pipeline batched through cuDF and PyTorch Lightning; replaced a GPR stall with cuML/RAFT (batch-scaling 150→50K+ @7.2ms/seq).
- ESM2
- cuDF
- cuML
- RAFT
- OmegaConf
Jun 2024 – Sep 2024
GenAI Engineering Intern @Talin Labs @Talin Labs
Deployed a fine-tuned Mistral-7B on Kubernetes with six-agent LangChain orchestration, 200ms@P95.
Deployed a fine-tuned Mistral-7B-Q8 on Kubernetes behind FastAPI with a six-agent LangChain orchestration. Sub-200 ms p95 for 10K users; 88% on 10K human-evaluated queries.
- Kubernetes
- FastAPI
- LangChain
- LangGraph

Web app
website-client@sushrutalgs.ai @sushrutalgs.ai
Web app for studying with the AI assistant, every answer backed by textbook citations. §case study
Web app for the sushrutalgs.ai platform where users study with the AI assistant, branching conversations into follow-ups with every answer backed by textbook citations, figures, and tables.
- Next.js 16
- React 19
- TypeScript
- Supabase
- Vercel
- case study
Retrieval backend
HybridFlow@sushrutalgs.ai @sushrutalgs.ai
Retrieves the most relevant medical knowledge and streams grounded answers to each question. §case study
Retrieval module for the sushrutalgs.ai platform that pulls the most relevant medical knowledge and generates grounded, streaming answers to each question.
- Python
- FastAPI
- Qdrant
- Neo4j
- case study
API gateway
backend-for-frontend@sushrutalgs.ai @sushrutalgs.ai
Gateway connecting the apps to the AI backend: sign-in, plan limits, streaming. §case study
Gateway module for the sushrutalgs.ai platform that connects the apps to the AI backend, signing users in, enforcing their plan limits, and streaming answers back. Live in production.
- TypeScript
- Hono
- Cloudflare Workers
- Supabase
- case study
iOS app
ios-app-client@sushrutalgs.ai @sushrutalgs.ai
Native iPhone app for studying with the AI assistant, with live streaming answers across devices. §case study
Native iPhone app for the sushrutalgs.ai platform where users study with the AI assistant, with answers that stream in live and sessions that carry across their devices.
- Swift 6
- SwiftUI
- supabase-swift
- Xcode Cloud
- case study
Data pipeline
Samhita@sushrutalgs.ai @sushrutalgs.ai
Turns surgical-textbook PDFs into a structured, searchable knowledge base for the platform. §case study
Data ingestion module for the sushrutalgs.ai platform that turns surgical-textbook PDFs into a structured, searchable knowledge base for the rest of the platform to draw on.
- Python
- BioLORD
- Cloudflare R2
- case study
Master's thesis
MACE-PINNpaper
Thesis oscillating neural network solver for stiff coupled physics equations. §case study
Master's thesis neural network that accurately solves the stiff, tightly coupled physics equations that standard single-network models struggle to capture.
- JAX
- Flax
- NumPy
- case study
- paper
ML platform
Yelp ML Platformgithub
End-to-end ML system on the Yelp dataset: recommendations and review sentiment from one API. §case study
End-to-end ML system on the full Yelp dataset that recommends businesses to users and reads the sentiment of their reviews, served from one fast API.
- PySpark
- FastAPI
- MLflow
- Docker
- NumPy
- case study
- github

2025
MS, Data Science @ASU @Arizona State University
High Performance Computing and Decision Analytics · GPA 3.71
- Statistical ML
- Distributed Systems
- GPU Computing
- Deep Learning Theory
- Numerical Methods
- Optimization
- Thesis: Coupled Ensemble Physics-Informed Neural NetworksThesis: Multi-Architecture Coupled Ensemble Physics-Informed Neural Networks
- GPA 3.71
2023
B.Tech, Computer Science @Nirma @Nirma University
Robotics, Computer Vision and Automation
- Algorithms
- Operating Systems
- Computer Networks
- Databases
- Compilers
- Linear Algebra
- Hazardous-asteroid classification with quantum MLClassification of potentially hazardous asteroids using supervised quantum machine learningIEEE Access, 2023
- MetaHate: hate-speech detection for metaverse gamingMetaHate: AI-based hate speech detection for secured online gaming in metaverse using blockchainSecurity and Privacy, 2024

Hi! I’m Rushir
Bhavsar.

Experience.

Founding Engineer @sushrutalgs.ai @sushrutalgs.ai

ML Engineer Intern @Cadence @OpenEye, Cadence Design Systems

GenAI Engineering Intern @Talin Labs @Talin Labs

Selected Projects.

website-client@sushrutalgs.ai @sushrutalgs.ai

HybridFlow@sushrutalgs.ai @sushrutalgs.ai

backend-for-frontend@sushrutalgs.ai @sushrutalgs.ai

ios-app-client@sushrutalgs.ai @sushrutalgs.ai

Samhita@sushrutalgs.ai @sushrutalgs.ai

MACE-PINNpaper

Yelp ML Platformgithub

Education.

MS, Data Science @ASU @Arizona State University

B.Tech, Computer Science @Nirma @Nirma University

Contact.

Hey! I'm looking for a role.

Hi! I’m Rushir Bhavsar.

Experience.

Selected Projects.

Education.

Contact.

Hi! I’m Rushir
Bhavsar.