Software developer

• Sole owner of the LCM module written in Typescript.
• LCM lies at the core of the product. It handles queueing, scheduling and execution of all major lifecycle processes.

• LCM prevents corruption of customer data, removes duplicate process, allows error handling, resumes process from where it left off to save time.
• Optimising memory usage of JS code and detecting memory leaks to reduce server cost.

Full Stack Developer & Machine Learning Engineer

• Built a project from scratch in Python that clusters similar user queries.
• Customer can add these clusters in a new intent.
• This improves chatbot performance and reduces the need for human intervention.
• Developed context-aware feature for the chatbot so it remembers the previous user input and responds based on that.
• Single-handedly built the most user requested feature Whatsapp integration.
• Converted handful of the existing APIs to GraphQL for effecient data retrieval.

Deep Learning Intern

• Trained and deployed language models using LSTM based Recurrent Neural Networks in Tensorflow. Improvement over traditional N-gram models.
• Splitting words in the training data to subwords which accuracy by 11%.
• Designed custom improved the deep learning model to predict conversation between two individuals.

Module lead

• Added a new module Holonomic functions in the open source Python library SymPy.
• This increases the number of functions that SymPy can integrate. successfully.
• Researched and developed algorithms for converting D-finite function to algaebric functions and vice versa.