• Machine Learning Algorithms - Prototype selection for K-NN neighbours
• Probabilistic Reasoning & Learning - EM algorithm to build a Recommender System, Logistic Regression to classify handwritten digits, Statistical Language Modelling using a mixture model
• Virtualization Techniques
• Algorithms in Computational Biology

• Co-authored a ​paper during the final year of college where different ML classification techniques ​were used to achieve the highest accuracy in binary classification of stance data. Was awarded the Silver Medal for the best final year project
• Explored ML & statistical methods in R to analyse fMRI images at Indian Institute of Technology, Delhi during the summer of 2015
• Created a web application where the user draws a weighted graph and the result is a minimum cost spanning tree based on Prim’s algorithm as part of the Algorithms course
• Built a conversational chat application that allowed group chats and file sharing

I worked for Goldman Sachs in their Global Markets Division, based out of Bengaluru. I've been a part of two automation projects and another project that involved revamping their Synthetic Client Reporting tech stack.

The first automation project involved helping traders across New York, London and Hong Kong automate booking trades on the company's proprietary software. My efforts were concentrated on designing, developing, optimizing and testing the workflow automation. I also devoted a significant part of my time in developing control scenarios in Python to alert in case of workflow failures. My second automation project involved setting up an automated regression environment that helped my team test and release their changes faster. This involved writing shell scripts and orchestrating the workflow.

Revamping Client Reporting tech stack involved automating batch workflows & scripts to extract reporting data based on synthetic stocks traded by clients on a given day and storing it in MongoDb. This also involved writing a Restful service in Java to expose reporting data extracted in MongoDb.

Right after college, I worked for American Express. I was the youngest and the newest member in a team of passionate developers that launched the company's Risk Management Platform called Smart Monitoring. I was responsible for training and deploying an ML model to predict and alert if an Amex’s debtor’s credit worthiness might be impacted based on articles being written about them in the news. I also wrote several RESTful microservices in Java using Spring Boot with HBase as the datasource.

Apart from this, I also built a CV parsing and ordering system. My team would receive hundreds of CVs for an open position and it would involve a lot of manual effort to go through all of them to identify the right candidates to interview. I used NLP processing techniques (LDA based NER, PoS Tagging, Named Entity Extraction, Topic Modelling etc.) to extract features from CVs and also trained a Naive Bayes based model that predicted how likely a candidate would convert the interview to an offer based on the data of candidates that were previously selected.