•
•
•
•
•
•
1
•
•
2
•
•
3
•
•
4
•
•
5
•
•
6
•
•
7
•
•
8
•
•
9
•
•
10
•
•
11
•
•
12
•
•
13
•
•
14
•
•
15
•
•
16
•
•
•
•
•
•
•
•
•
•
1
•
•
2
•
•
3
•
•
4
•
•
5
•
•
6
•
•
7
•
•
8
•
•
9
•
•
10
•
•
11
•
•
12
•
•
13
•
•
14
•
•
15
•
•
16
•
•
•
•
•
•
•
•
1
•
•
2
•
•
3
•
•
4
•
•
5
•
•
6
•
•
7
•
•
8
•
•
9
•
•
10
•
•
11
•
•
12
•
•
13
•
•
14
•
•
15
•
•
16
•
•
•
•
•
•
•
•
•
•
1
•
•
2
•
•
3
•
•
4
•
•
5
•
•
6
•
•
7
•
•
8
•
•
9
•
•
10
•
•
11
•
•
12
•
•
13
•
•
14
•
•
15
•
•
16
•
•
17
•
•
18
•
•
19
•
•
20
•
•
21
•
•
22
•
•
23
•
•
24
•
•
25
•
•
•
•
•
•
•
•
•
•
•
1
•
•
2
•
•
3
•
•
4
•
•
5
•
•
6
•
•
7
•
•
8
•
•
9
•
•
10
•
•
11
•
•
12
•
•
13
•
•
14
•
•
15
•
•
16
•
•
17
•
•
18
•
•
19
•
•
20
•
•
21
•
•
22
•
•
23
•
•
24
•
•
25
•
•
26
•
•
27
•
•
28
•
•
29
•
•
30
•
•
•
•
•
•
My work experience
Waters Corporation
GenAI and ML Intern
Jun 2025 - Aug 2025
Internship
Implemented LLM evaluation framework for AI search Engine automating multi-dimensional RAGAS and LLM-as-a-Judge testing
Synthetically generated 2K+ test samples with Claude Sonnet 4 via AWS Bedrock across 3 query types, cutting SME review time by 50%
Integrated retrieval pipelines (OpenSearch + S3) to streamline 400k+ chunks, optimizing resource & scaling by 2.5x across teams
Ensured human-in-the-loop automation with WebSocket-based Streamlit dashboard for SMEs to track metrics and drive AI improvements
Closed 24 story points in one sprint and live-demoed an internal feature to 200+ employees across US/India, doubling team productivity.
Waters Corporation
GenAI and ML Intern
Jun 2025 - Aug 2025
Internship
Implemented LLM evaluation framework for AI search Engine automating multi-dimensional RAGAS and LLM-as-a-Judge testing
Synthetically generated 2K+ test samples with Claude Sonnet 4 via AWS Bedrock across 3 query types, cutting SME review time by 50%
Integrated retrieval pipelines (OpenSearch + S3) to streamline 400k+ chunks, optimizing resource & scaling by 2.5x across teams
Ensured human-in-the-loop automation with WebSocket-based Streamlit dashboard for SMEs to track metrics and drive AI improvements
Closed 24 story points in one sprint and live-demoed an internal feature to 200+ employees across US/India, doubling team productivity.
Waters Corporation
GenAI and ML Intern
Jun 2025 - Aug 2025
Internship
Implemented LLM evaluation framework for AI search Engine automating multi-dimensional RAGAS and LLM-as-a-Judge testing
Synthetically generated 2K+ test samples with Claude Sonnet 4 via AWS Bedrock across 3 query types, cutting SME review time by 50%
Integrated retrieval pipelines (OpenSearch + S3) to streamline 400k+ chunks, optimizing resource & scaling by 2.5x across teams
Ensured human-in-the-loop automation with WebSocket-based Streamlit dashboard for SMEs to track metrics and drive AI improvements
Closed 24 story points in one sprint and live-demoed an internal feature to 200+ employees across US/India, doubling team productivity.
Edlight, PBC
Machine Learning Intern
Sep 2024 - Dec 2024
Internship
Delivered multi-agent chatbot with guardrails using LangGraph and FastAPI, with streaming RESTful API endpoints on AWS Lambda
Achieved 180ms p95 TTFT via dynamic agent routing: SQL for curriculum, RAG for resources, direct responses for general questions
Architected SQL-Agent with custom tools to dynamically fetch data from multiple PostgreSQL tables using natural-language queries
Improved ChromaDB retrieval using CRAG and reranker, increasing faithfulness (+13.4%) and reducing hallucinations (-32.5%)
Optimized multi-modal API with queuing and threading to reduce latency by 80% while sustaining 30K+ monthly requests.
Built an assignment image to IEP goals matching system utilizing a multi-step Qwen2-VL-7B-Instruct model that won week-long hackathon and became the biggest product release for the company in Q1 2025
Edlight, PBC
Machine Learning Intern
Sep 2024 - Dec 2024
Internship
Delivered multi-agent chatbot with guardrails using LangGraph and FastAPI, with streaming RESTful API endpoints on AWS Lambda
Achieved 180ms p95 TTFT via dynamic agent routing: SQL for curriculum, RAG for resources, direct responses for general questions
Architected SQL-Agent with custom tools to dynamically fetch data from multiple PostgreSQL tables using natural-language queries
Improved ChromaDB retrieval using CRAG and reranker, increasing faithfulness (+13.4%) and reducing hallucinations (-32.5%)
Optimized multi-modal API with queuing and threading to reduce latency by 80% while sustaining 30K+ monthly requests.
Built an assignment image to IEP goals matching system utilizing a multi-step Qwen2-VL-7B-Instruct model that won week-long hackathon and became the biggest product release for the company in Q1 2025
Edlight, PBC
Machine Learning Intern
Sep 2024 - Dec 2024
Internship
Delivered multi-agent chatbot with guardrails using LangGraph and FastAPI, with streaming RESTful API endpoints on AWS Lambda
Achieved 180ms p95 TTFT via dynamic agent routing: SQL for curriculum, RAG for resources, direct responses for general questions
Architected SQL-Agent with custom tools to dynamically fetch data from multiple PostgreSQL tables using natural-language queries
Improved ChromaDB retrieval using CRAG and reranker, increasing faithfulness (+13.4%) and reducing hallucinations (-32.5%)
Optimized multi-modal API with queuing and threading to reduce latency by 80% while sustaining 30K+ monthly requests.
Built an assignment image to IEP goals matching system utilizing a multi-step Qwen2-VL-7B-Instruct model that won week-long hackathon and became the biggest product release for the company in Q1 2025
Impacter AI
AI Developer Intern
Sep 2024 - Dec 2024
Internship
Built ‘SalesPal’, AI-powered real-estate sales agent with Azure OpenAI, CosmosDB, and Databricks to automate personalized client pitching
Integrated CrewAI-based email and call agents (SendGrid, Twilio, MCP), automating contextual outreach to 1K+ concurrent users
Wrapped SalesPal in a scalable Flask web app, containerized with Docker, deployed on Azure Kubernetes Service with CI/CD pipelines
Impacter AI
AI Developer Intern
Sep 2024 - Dec 2024
Internship
Built ‘SalesPal’, AI-powered real-estate sales agent with Azure OpenAI, CosmosDB, and Databricks to automate personalized client pitching
Integrated CrewAI-based email and call agents (SendGrid, Twilio, MCP), automating contextual outreach to 1K+ concurrent users
Wrapped SalesPal in a scalable Flask web app, containerized with Docker, deployed on Azure Kubernetes Service with CI/CD pipelines
Impacter AI
AI Developer Intern
Sep 2024 - Dec 2024
Internship
Built ‘SalesPal’, AI-powered real-estate sales agent with Azure OpenAI, CosmosDB, and Databricks to automate personalized client pitching
Integrated CrewAI-based email and call agents (SendGrid, Twilio, MCP), automating contextual outreach to 1K+ concurrent users
Wrapped SalesPal in a scalable Flask web app, containerized with Docker, deployed on Azure Kubernetes Service with CI/CD pipelines
Hyperlab Sportech
Founding ML Engineer
Jan 2022 - Aug 2023
Full-Time
Launched a ML-driven iOS app using Swift with 50K+ downloads and 20K+ DAU, secured $100K funding via Shark Tank and angel investors
Built multimodal AI Coach (LangChain, GPT-4, Llama 3, AstraDB) for vision and text processing with FastAPI backed inference
Created AstraDB vector store from 30K+ files via semantic chunking and Fine-Tuned Llama 3 to reduce hallucinations by 60%
Developed ETL pipelines using AWS Glue, Spark, and Kinesis to process 10TB+ of athlete performance data and insights
Developed CNN-based Table Tennis ball prediction system, trained on AWS SageMaker using PyTorch and unified data warehouse
Performed Multi-GPU Distributed Training of LSTM model with Ray on 4 Nvidia GPUs, containerized and deployed on AWS EC2
Hyperlab Sportech
Founding ML Engineer
Jan 2022 - Aug 2023
Full-Time
Launched a ML-driven iOS app using Swift with 50K+ downloads and 20K+ DAU, secured $100K funding via Shark Tank and angel investors
Built multimodal AI Coach (LangChain, GPT-4, Llama 3, AstraDB) for vision and text processing with FastAPI backed inference
Created AstraDB vector store from 30K+ files via semantic chunking and Fine-Tuned Llama 3 to reduce hallucinations by 60%
Developed ETL pipelines using AWS Glue, Spark, and Kinesis to process 10TB+ of athlete performance data and insights
Developed CNN-based Table Tennis ball prediction system, trained on AWS SageMaker using PyTorch and unified data warehouse
Performed Multi-GPU Distributed Training of LSTM model with Ray on 4 Nvidia GPUs, containerized and deployed on AWS EC2
Hyperlab Sportech
Founding ML Engineer
Jan 2022 - Aug 2023
Full-Time
Launched a ML-driven iOS app using Swift with 50K+ downloads and 20K+ DAU, secured $100K funding via Shark Tank and angel investors
Built multimodal AI Coach (LangChain, GPT-4, Llama 3, AstraDB) for vision and text processing with FastAPI backed inference
Created AstraDB vector store from 30K+ files via semantic chunking and Fine-Tuned Llama 3 to reduce hallucinations by 60%
Developed ETL pipelines using AWS Glue, Spark, and Kinesis to process 10TB+ of athlete performance data and insights
Developed CNN-based Table Tennis ball prediction system, trained on AWS SageMaker using PyTorch and unified data warehouse
Performed Multi-GPU Distributed Training of LSTM model with Ray on 4 Nvidia GPUs, containerized and deployed on AWS EC2
Nirma University
Student Researcher
Jan 2022 - May 2022
Extracurricular
Curated a dataset of a Li-ion battery pack encompassing over 150k rows, extracting actionable insights for data-driven strategies.
Utilized Random Forest regression model to predict optimal temperature during charging/discharging, improving battery health by 12%.
Employed clustering techniques to understand current rates and temperature of the battery pack over 1300 charge-discharge cycles
Nirma University
Student Researcher
Jan 2022 - May 2022
Extracurricular
Curated a dataset of a Li-ion battery pack encompassing over 150k rows, extracting actionable insights for data-driven strategies.
Utilized Random Forest regression model to predict optimal temperature during charging/discharging, improving battery health by 12%.
Employed clustering techniques to understand current rates and temperature of the battery pack over 1300 charge-discharge cycles
Nirma University
Student Researcher
Jan 2022 - May 2022
Extracurricular
Curated a dataset of a Li-ion battery pack encompassing over 150k rows, extracting actionable insights for data-driven strategies.
Utilized Random Forest regression model to predict optimal temperature during charging/discharging, improving battery health by 12%.
Employed clustering techniques to understand current rates and temperature of the battery pack over 1300 charge-discharge cycles
Team Stallions
Captain
Jan 2021 - Dec 2021
Extracurricular
Trained and Managed a team of 40 students, while leading the development of 'Mercury,' an Electric Solar Vehicle with 150km Range and 60kmph top speed.
Secured a decisive victory in Asia's Largest Electric Solar Vehicle Competition 2021, outperforming competitors by over 500 points and earning 8 out of 11 trophies.
Engineered a patented steering system for Vehicle's Autonomous System, a pivotal contribution that led to winning the 'Best Autonomous Round’ Award
Team Stallions
Captain
Jan 2021 - Dec 2021
Extracurricular
Trained and Managed a team of 40 students, while leading the development of 'Mercury,' an Electric Solar Vehicle with 150km Range and 60kmph top speed.
Secured a decisive victory in Asia's Largest Electric Solar Vehicle Competition 2021, outperforming competitors by over 500 points and earning 8 out of 11 trophies.
Engineered a patented steering system for Vehicle's Autonomous System, a pivotal contribution that led to winning the 'Best Autonomous Round’ Award
Team Stallions
Captain
Jan 2021 - Dec 2021
Extracurricular
Trained and Managed a team of 40 students, while leading the development of 'Mercury,' an Electric Solar Vehicle with 150km Range and 60kmph top speed.
Secured a decisive victory in Asia's Largest Electric Solar Vehicle Competition 2021, outperforming competitors by over 500 points and earning 8 out of 11 trophies.
Engineered a patented steering system for Vehicle's Autonomous System, a pivotal contribution that led to winning the 'Best Autonomous Round’ Award
Get in touch with me at
henil.gajjar2102@gmail.com
42.3555° N, 71.0565° W
Thursday, Nov 13, 2025
Get in touch with me at
henil.gajjar2102@gmail.com
42.3555° N, 71.0565° W
Thursday, Nov 13, 2025
Get in touch with me at
henil.gajjar2102@gmail.com
42.3555° N, 71.0565° W
Thursday, Nov 13, 2025