* Design and Implement the back-end pipeline in FastAPI* Apply image processing methods and train computer vision models with TensorFlow to identify particular activities along with tracking analytics* Discover relevant metrics and build dashboard prototypes* Automate the end-to-end AI pipeline from data gathering and training to deployment* Fine tune Large Language Models for text generation and summarization use cases with chat like experience* Setup infrastructure for deploying AI models in production

- Design pipelines for cleaning, storing and versioning scraped data- Trained, fine-tuned and tested a Large Language Model (LLM) to understand context, generate / execute bash scripts as a background service and translate output to formatted texts- Built Object Detection Models for Optical Character Recognition (OCR) allowing us to extract relevant information from position and contents of characters detected- Create a simple UI in Vanilla JS and HTML to test models and visualize results appropriately

- Designed the backend infrastructure of Lolly from scratch in FastAPI.- Integrated third party payment applications such as Stripe, Paypal, Shopify, BigCommerce, Google, Facebook, and Tiktok.- Built a chat system along with reporting service for retrieving daily analytics form social media platforms.- Designed a notification service for regular user activites occurring on the platform.- Created optimised queries for gathering data from Mongo, SQL and Elasticsearch databaes.- Deployed and monitored all the above mentioned services successfully in production.

- Utilised APIs provided from Facebook and Google to build Ad Campaigns programtically with Flask framework.- Built ML models to optimize Ad Keywords and Negative Keywords on user campaigns.

- Analyzed Covid-19 cases across several regions and built time series models for predicting the number of cases in the upcoming weeks.- Created and demonstrated a general purpose end-to-end machine learning pipeline from data loading and preprocessing to model building training and testing.- Made APIs for easy data capture and data flow from source to various NoSQL databases.

- Built convolutional neural networks for object detection such as lane markings, traffic lights and stop signs.- Designed, trained and tested deep learning models such as RNN along with Reinforcement Learning for autonomous vehicles in controlled environments.