1. A platform for users to create chatbots using documents and website content using Rafiq AI.
2. A chatbot for US citizens to get updated information of their respective townships. Implemented the whole
backend and LLM integration: Munichat.


ML Engineer

<p>1. Shopadvisor AI: Built an AI shopping advisor for e-commerce companies using LLMs.</p><p>2. Lead everything in product and engineering.</p>

Founding AI Engineer

1. Built a no-code ML platform for users to build, train and deploy their ML models.
2. Created all kinds of pipelines in the image and text domain and connected these services with AWS and Runpod
for deployment.
3. Deployed ML models on serverless Runpod platform as inference APIs.
4. Deployed zero-shot models for labeling tasks such as image object detection and segmentation.
5. Added few-shot pipelines for text-related tasks to enable training models with a small amount of data.

ML Lead

1. Muse App: Deployed a text-to-image generation model (stable-diffusion) in production, which handles thousands of concurrent users. Implemented a load balancer with dynamic batching to improve the model serving performance from 10 to 500 concurrent users using micro-services. On the launch, the App hit around 8000 requests in 2 days without any failures.
2. Led the stable diffusion research with 2 team members, where we explored the limitation of the Lightning framework when a foundational model is deployed in production.
3. Collaborated with the Colossal AI team to integrate the Colossal AI engine that implements different parallelism algorithms that are especially interesting for developing SOTA transformer models.
4. Additional notable contributions were Tuner callbacks, FSDP auto-wrappers, and DeepSpeed integration improvements.
5. The project I worked on and maintained is lightning which is open-source and currently has 20k stars and 7510+ contributors. By the time I joined, I was already in the top 10 as a core contributor, and now I am in the top-5. It is built on top of PyTorch from Meta.
6. Advise on, assist, and support the implementation, optimization, and development of those improvements and new technologies in the framework, along with timely releases.
7. Integration and maintenance of SOTA deep-learning optimization techniques developed by research labs into the framework.
8. Ensured proper maintenance of documentation that is completely user-facing by ensuring everything is updated and explained with real-world examples.
9. Ensured continuous community engagement and discussions that helped increase the reach of the framework and helped in its development.


Research Engineer

1. Slashed the processing time from days to an hour by building a complete automated pipeline to generate patient profiles for HRA along with model deployment on AWS and GitHub CI/CD pipeline.
2. Collaborated in the in-house development of the Datalake warehouse using PySpark. This helped to ensure data integrity and correctness in various verticals of business where it is being used.
3. Analyzed the HRA and Telehealth services thus reducing the cost incurred, operation optimization, and managed the planning & deployment of design and procedures for weekly metric reports. The analysis helped in taking major business decisions to improve the processes. Lead the team with 2 interns.
4. Improved the contact rate for OCs by 30% to non-approachable customers by developing an automated pipeline to mine historical data and find relevant contact information for upcoming/ongoing projects.
5. Built a machine learning model that can predict the outreach consultant headcount for an incoming project.

Data Scientist

I have been working with python for more than 3 years and have build data analytics and machine learning projects.

I have worked on a data analytics projects at work and have created ML pipelines at my day jobs. I was a research engineer at Lighning AI and have build PyTorch Lightning open-source projects which went to 25k stars and became top-5th contributor there.

Personal details

Rohit G.

Summary

Technical skills

Work Experience

Education