Feb 2023 – Dec 2024 · 1 yr 10 mos · Bengaluru, Karnataka, India · On-site

• Building credit risk models
• Developing LLM based models to support customer experience

Feb 2019 – Jan 2023 · 3 yrs 11 mos · Bengaluru, Karnataka, India

• 1. Working on optimal assignment of Delivery Partners to orders so that overall customer experience is maximized with minimal cost of delivery. Formulated this problem as 0-1 ILP(Integer Linear Programming) with multi objective and solved this Operational Research problem using Gurobi.
• 2. Built Restaurants Listing model based on efficiency of the restaurant and conversion rate of the customer.
• 3. Working on formulating Assignment and batching as a Reinforcement Learning problem.

Jul 2017 – Jan 2019 · 1 yr 6 mos · Bengaluru Area, India

• 1.Automated Financial Crime Insights by extracting information using NLP techniques like Named Entity Recognition(NER), Coreference Resolution, Sentiment Analysis.
• 2.Trained a binary classifier on Adverse Press data using Ensembled Learning on Naive Bayes, Random Forest, SVM where TF-IDF, Word2Vec, GloVe were used for feature representation.
• 3.Optimized processing time of neural-coref (a coreference library) from 57 sec to 3 sec.
• 4.Developed a unsupervised document summarization system using Word2Vec and DBSCAN.
• 5.Improved accuracy of NER module from 85.85% to 89%(PERSON) and 86%(ORGANISATION).
• 6.Designed and Developed a software application to simplify collection of prediction feedback.

May 2016 – Jun 2016 · 1 mo · Bengaluru Area, India

• 1.Formulated a method to auto-suggest the succeeding Tizen(OS used by Samsung) API for code editor.
• 2.Used SimHash for hashing and BK tree data structure to get approximate string matching.
• 3.Bayesian Networks was employed to generate the frequent usage pattern.