Software, Machine Learning, Data

^

My expertise intersects

Hero Image

Projects

View some of my latest projects

Project
GeoSpatial Data API Service
Docker
FastAPI
Airflow
Snowflake
AWS
Algorithm Design
Data as a Service
AWS
CI/CD

Cloud-based API Service for GeoSpatial Data with Authentication, User Plans(API Call Limits), and Dashboards

Project
Predicting Cab Fare for Lyft and Uber Rides: A Data-Driven Approach
Linear Regression
Decision Tree
Random Forest
Gradient Boosting Regressor
Machine Learning

Enhancing cab ride experience with machine learning. Predicting Lyft and Uber fares based on weather, time and day of week for better customer satisfaction.

Project
Meeting Intelligence with ChatGPT and Whisper API
Large Language Models (LLM)
Retrieval Augmented Generation
Airflow
AWS
Algorithm Design
Transform, Load (ETL)
Model as a Service
OPEN AI API
AWS
Data Engineering

Chatbot to analyze and converse against Zoom/Teams meetings transcripts

Project
Project Vending Machine
Machine Learning
Docker
FastAPI
Airflow
Algorithm Design
Model as a Service
OPEN AI API
CI/CD
Github API
MongoDB

Intelligent boiler plate code generator for any given data engineering project in under 4 minutes based on a prompt and return a github repository with the code

Project
Heart Monitor with Polar H10 Sensor
Medical Devices
IEC 62304
Software Development
Bluetooth Low Energy

- Measures Heart Rate with an accuracy of ~97% accuracy. - Currently Working on adding an ECG recorder to the monitor.

Project
LLM Powered HealthCare Chatbot using Neo4j RAG and Knowledge Graph
Knowledge Graphs
Large Language Models (LLM)
Retrieval Augmented Generation
Named Entity Recognition (NER)
spaCy

- Built a Patient Context aware Chatbot that can respond to queries with a 100% accuracy. - Implemented a RAG and Knowledge Graph using Neo4j

Project
Chest Cancer Classification using CNN and transfer learning with VGG16
Deep Learning
MLOps
Tensorflow
Keras
Docker
AWS ECR
Flask
Transfer Learning
CNN
Convoluted Neural Networks
DVC

Developed deep learning model for chest cancer classification integrating MLOps tools; achieved 92% accuracy; deployed on AWS.