Venkatesh Banginwar

Energetic and team player, Senior Associate Consultant with exceptional and demonstrable quantitative abilities to perform rigorous and creative analyses, learn new industries quickly and function as integrated member of project team. Manages and delivers assigned responsibilities within scope and budget.

KNOWLEDGE REPOSITORY CHATBOT

• A chatbot catering to internal requirement which channelizes information gathered from transcripts of internal meeting/conference videos which were generated using 'WHISPER', then vectorized and stored in vectorDB, a RAG Pipeline was created by using LangChain which gave the expected output, GPT-4 was used to create the embeddings. A slick UI was developed with help of Streamlit., Worked on implementing RAG Pipeline., Creating embeddings using GPT-4 and prompt tuning to get better result., Extracting transcripts from videos and storing them in vectorDB., Contributed in creating a interactive UI with help of Streamlit.

JAVA TO PYTHON CODE CONVERSION

• A pipeline was created where user can upload a zip folder containing java files and the model will generate respective python files which can be downloaded as a zip folder. UI was created using the powerful Streamlit library of Python, Worked on creating a parsing mechanism for input java code using tree-sitter., Created one-to-one file translation logic and functionality to read files from zip folder., Prompt tuning to get better output from the model., Contributed in creating a interactive UI with help of Streamlit.

 ML PALTFORM FOR COMPETITION 

• A ML platform was created for the employees of the organization where one can start a ML competition by creating a use-case by uploading dataset and others can upload their own models, which after evaluation will be ranked on a leaderboard according to there performance. The project was developed by implementing microservice architecture., Created multiple python services to handle data and create a statistics for dataset., Implemented confluent-kafka system to handle messages from various different services., Designed API's with help of flask-restplus and pymongo., To maintain concurrency implemented Thespian actor system.

Air Side Park

• A web application was developed for a parking aggregator from where their customers will be able to book a parking space for their car for a desired time period, the app will then inform the driver to pick up the car at the time of customers departure and drop it back when the customer arrives., Created a database model to store data of entire app., Designed database API's to transact required data for the various needs., Developed API's to integrate frontend., Wrote a script in Django Celery to read a mail box and extract excel attachments periodically.