About
Highly analytical Data Scientist with a strong foundation in fraud detection, risk analytics, and secure ML systems, evidenced by boosting BERT classification accuracy by 18% and reducing ML pipeline latency by 70%. Proven expertise in A/B testing, campaign evaluation, and developing robust dashboards, enabling data-driven policy decisions and significant operational risk reduction across diverse industries.
Work
Remote, Remote, US
→
Summary
As a Data Science Intern, I leveraged advanced ML techniques to enhance classification accuracy and optimize data pipelines for a leading AI hiring startup, directly contributing to improved operational efficiency and data visibility.
Highlights
Boosted BERT classification accuracy by 18% across 50K+ job listings and 30+ categories, enhancing job matching precision.
Cut ML pipeline latency by 70% via DAG refactoring and SQL caching optimizations, significantly improving data processing speed.
Reconciled data mismatches between source and Tableau dashboards for accurate reporting, ensuring data integrity and reliability for key metrics.
Built self-serve Tableau reports for non-technical users, tracking pipeline and conversion KPIs to empower data-driven decisions.
Hyderabad, Telangana, India
→
Summary
As a Research Assistant, I developed a high-accuracy ensemble-based ML model for crop recommendation and co-authored an IEEE paper, contributing to agricultural innovation and farmer support.
Highlights
Developed an ensemble-based crop recommendation model achieving 98.9% accuracy using 300K+ records, enhancing agricultural productivity.
Co-authored an IEEE paper and deployed a Flask-based cloud tool for real-time crop recommendations to farmers, disseminating research into practical applications.
Bengaluru, Karnataka, India
→
Summary
As a Data Scientist-NLP & GenAI Intern, I developed and deployed AI-powered solutions, including RAG chatbots and predictive features, to enhance financial and healthcare query resolution and optimize user engagement for an AI product team.
Highlights
Deployed a RAG-powered chatbot, resolving 65% of domain-specific finance/health queries and improving user support efficiency.
Conducted A/B testing to optimize preventive campaign reach and reduce user drop-off, enhancing user engagement and retention.
Designed predictive features instrumental in fraud/eligibility classification dashboards, bolstering system accuracy and security.
Collaborated with non-technical stakeholders to translate complex predictive insights into actionable outreach strategies, bridging technical and business objectives.
Education
Skills
Languages
Python, SQL, R, SAS, Bash, C++.
ML & Risk Modeling
scikit-learn, XGBoost, BERT, A/B Testing, Anomaly Detection, Time Series.
Tools & Infra
Tableau, Power BI, MLflow, Airflow, FastAPI, GitHub Actions.
Data & Cloud
BigQuery, Redshift, AWS (S3, Lambda), Azure (ADF, Blob), Hadoop.
NLP & GenAI
Hugging Face, Transformers, RAG, FAISS, LangChain.
Risk Practices
Fraud detection, Conversion funnel optimization, Lifecycle analytics, Compliance dashboards.