Jul 2023 – Oct 2025 · 2 yrs 3 mos · Hyderabad, Telangana, India · On-site

Jun 2022 – Jul 2023 · 1 yr 1 mo · Bengaluru, Karnataka, India · On-site

• Worked on NLP tasks like Multilingal Multi-label Text Classification, De-duplication and Sentiment Analysis to improve data relevancy on platform
• Reduced model's complexity and size by 50% without significant loss in performance, using Knowledge Distillation technique
• Leveraged Optimum ONNX Runtime and DeepSparse Quantization techniques to reduce existing ML model’s Inferencing and Training time by 75%
• Proposed and migrated all DS services to a single GKE cluster and integrated multiple Monolingual models into one model resulting in huge infra-cost savings

May 2021 – Jul 2021 · 2 mos · WFH

• Built a prototype that increased the throughput in the existing 5G Resource Schedulers using ML Algorithms.
• Read multiple Research Papers and Literature Surveys, shortlisted some, gave presentations, and finally
• implemented one.
• Generated dataset using Simulink, and used various Supervised learning models and Time Series model
• Got exposure to ML-Answers(MathWorks Community Page), MATLAB, its toolboxes like Classification Learner,
• Regression Learner, Neural Net Time Series, Optimization, etc.
• Reported highest number of gecks (bug/enhancement on the new Release product) among interns at Hyderabad

Jul 2019 – Aug 2020 · 1 yr 1 mo · Pune Area, India

Jan 2019 – Jul 2019 · 6 mos · Mumbai Area, India

• Currently working for Eugenie, a startup within Fractal which is an AI-backed algorithmic suite that helps businesses spot, explore and exploit business anomalies that have bottom-line impact. Eugenie’s state-of-the-art machine learning algorithms finds out relevant actionable insights for business along with root cause and possible diagnostics across domains and functions.
• Objective is to develop the automated insight discovery capabilities of Eugenie.
• Responsible for building the product, developing new algorithms, improving the accuracy and speed of the existing algorithms and optimising the overall product.
• Working on a ML framework to analyze big data and generate fast and accurate results.
• Writing a time-series based forecasting engine on large scale data using ARIMA, FbProphet, Holt Winter's models in Python and Apache Spark.

Jun 2018 – Aug 2018 · 2 mos · Bengaluru Area, India

• "Tell me I forget, Teach me I remember, Involve me I learn” – Xun Kuang.
• 1Learn app assists the student in a smart way with interactive bite-sized content made for today’s generation.
• I was responsible for building a platform where you just place the app on the textbook, it will recognize what concept it is and it will show 1 minute animated video, summary flashcards and questions curated to the particular concept.
• Also built YouTube Video Recommendation Engine for students to micromanage their time and effort in searching the best video online.
• Worked with Google's Vision and YouTube Data API.
• Had hands-on experience with NLP and its libraries like Mallet, Spacy, nltk, TextBlob.
• Worked on Text Classification and Categorization, tf_idf, n-grams, POS Tagging, used models like Naive-Bayes Classifier, Support Vector Machine(SVM).
• First-hand experience of working in a challenging and unpredictable environment of a start-up. Got exposed to a culture of managing deliverables in a stipulated time
• Explored peer-peer learning opportunities at IIM-B and got exposed to a holistic work-learning ecosystem