"I love art. I love data. How could I not love the art of data science?"
I'm a passionate data science enhusiast always on the hunt for the next exciting problem to solve. I aspire to a role that fulfills my innate desire to solve impactful problems of the world. I work hard and don't stop till I achieve what I've set out to. I'm on a constant journey of self-improvement, and firmly beleive in adding value to the team I'm a part of.
If that sounds like something that interests you, shoot me an email at [email protected] or connect with me on LinkedIn, and maybe we can solve the next big problem together!
- M.S, CS | University of Southern California, Los Angeles (Dec 2024)
- B.Tech, CSE | Vellore Institute of Technology, Vellore (Sept 2020)
Business Systems Analyst @ Wolters Kluwer (Jan 2020 - Dec 2022)
- Created an interactive analytics tool in Splunk that mined real-time app logs and described 15+ use-cases
- Diagnosed production issues and aided incident management using Splunk and MySQL over 28 monthly sprints
- Defined detailed requirements for 25+ use-cases as a product specialist, serving 58 US jurisdictions
- Partnered with solution architects to design a multi-cloud DR strategy in AWS/Azure, impacting 74k users
- Crafted the integration of virus detection workflow into a web application, achieving 35 uploaded-file-scans/minute
- Revised 35+ use-cases for a web app to sustain an influx of 115k new users through an extensive GAP analysis
Data Science Intern @ Mphasis Limited (May 2019 - June 2019)
- Conducted extensive research on process mining in R, experimenting on 500K rows of in-patient activity data
- Proposed Graph DB as an effective store for join-intensive activity data, to accelerate response times by 50%
- Implemented a POC in the NeO4j GraphDB, generating mined process graph and improving query times by 20%
- Optimized query performance using CypherQL to improve 15 methods of unearthing KPIs, enhancing readability
- Refined process graph visualization using vis.js with 3 adaptive KPI indicators and animated edge traversal
- Applied Random Forest on healthcare data to predict users’ next activity in the journey with an accuracy of 70%
- Fine tuned GPT 2 and LLAMA 2 using Low Rank Adaption (LoRA) on a dataset of 200K Medium Articles
- Devised a weighted scoring mechanism, to assess coherence, tone, accuracy and writing style on a 10-pt scale
- Experimented with prompting techniques to outscore baseline models by 15% across 40 generated articles
- Crafted a BeautifulSoup web-scraper for automated IPL 2023 data collection from 10 teams and 74 matches
- Processed, cleaned, and transformed data using Pandas and identified 8 KPIs to evaluate player performance
- Curated Best XI from player stats using 8 extracted KPIs, showcased on an interactive Tableau dashboard
- Trained a 3-layer CNN + MLP architecture on 214 HiRISE subframe images (5120*5120 px) of Martian terrain
- Deployed various data augmentation and processing techniques to achieve a test macro avg f1-score of 92%
- Built a model to process skin lesion images and determine if they are cancerous by evaluating physical characteristics
- Used median filtering, clustering, edge detection techniques in MATLAB to deduce results with 92% accuracy
- Analyzed, cleaned, and pre-processed 2,000,000 records of Amazon customer review data in Python to prepare sample dataset
- Performed binary sentiment classification, achieving an accuracy of 89% by fine-tuning machine learning models
- Performed time series analysis on 16 years of PJM Interconnection East region Energy Consumption data measured in MW
- Extracted features & calculated importance using an XGBoost model. Predicted future energy consumption with 86% accuracy
- Conducted a comprehensive GUI evaluation of an interactive game developed in Python using Gesture Recognition techniques
- Evaluated Nielsen’s heuristics and conducted trials across age groups to study cognitive ability based on game performance
- Rewarded scholarship across 4 years of UG study in recognition of VITEEE’16 rank (top 3.5%) and excellent academic performance
- Received the prestigious enterprise-wide Chairman's Award for digital transformation of a critical application during the pandemic
- Python, R, JavaScript, SQL, CypherQL, MATLAB, HTML/CSS, Search Processing Language
- Pandas, NumPy, Matplotlib, Seaborn, Scikit-Learn, NLTK, Torch, TensorFlow, Image
- PyTorch, AWS, Tableau, Splunk, Git, Huggingface, VS Code, JIRA, LucidChart
- MySQL, PostgreSQL, NeO4j, Snowflake