Data Scientist and AI Engineer skilled in machine learning, NLP, data analytics, Gen AI seeking to apply technical expertise in AI-driven problem solving and automation to deliver impactful business outcomes within a forward-thinking organization.
Overview
2
2
years of professional experience
1
1
Certification
Work History
AI/ML Engineer
HCL Tech
Bengaluru
11.2024 - Current
Building, training, and deploying supervised machine learning and NLP-based classification models to automate workflows.
Architecting multi-stage classification frameworks that are minimizing manual intervention and driving down operational costs.
Driving model lifecycles, covering data preprocessing, feature engineering, training, hyperparameter optimization, and validation.
Applying advanced NLP methodologies such as TF-IDF vectorization, text classification, and transformer architectures (BERT).
Establishing evaluation frameworks leveraging metrics like precision, recall, PSI, KS statistic, capture rate, MAE, rogue score, and WMAPE.
Collaborated with cross-functional teams to automate call classification and email routing through AI-based systems, enhancing accuracy and efficiency.
Refining protected models by incorporating NLP pipelines that identify and filter abusive or sensitive communication.
Creating GenAI-driven POCs using LangChain and prompt routing for project management automation.
Documenting all development, validation, and governance activities.
Automated manual processes, generating estimated cost savings of $670,000 through implementation of advanced predictive models.
Achieved 85% accuracy and 83.5% recall for 2-stage classification model, reflecting strong model performance and reliability.
Created a dropout prediction model during the internship that achieved an accuracy rate of 85%, driving a 15% reduction in dropout rates.
Enhanced the average call handling time model using SARIMAX, achieving a WMAPE value of 6.5%, improving operational efficiency.
Data Science intern(paid)
Learnbay
Bengaluru
01.2024 - 10.2024
Designed a student dropout prediction model with 85% accuracy rate, enabling engagement strategies and reducing dropout rates by 15%.
Collaborated with academic advisors to generate actionable insights.
Created a time-series model for analyzing average call handling time, applying SARIMAX methodology, achieving WMAPE of 6.5%.
Education
Bachelor of Engineering - Mechanical
SJBIT
Bengaluru
01.2020
Skills
Python
SQL
Numpy
Pandas
Langchain
Ilama Index
Langgraph
Crew AI
Power BI
Snowflake
Machine Learning
Predictive Modeling
Natural Learning Processing
Time Series Forecasting
Data Visualization/ BI Reporting
Python Programming
Statistical Analysis
Supervised/ Unsupervised Learning
Model Evaluation
Data Cleaning & Preprocessing
Prompt/ Feature Engineering
Automation Frameworks
SQL Querying & Data Extraction
AI Model Monitoring/ Governance
Cross-functional Collaboration
Certification
Data Science and Artificial Intelligence Certificate, IBM
Python Certificate, IBM
Machine Learning Certificate, IBM
Machine Learning in Production
Microsoft Certified: Azure Data Fundamentals
Partner Training: Gen AI & LLM on Databricks
Projects
Mail Classification Model, Python, NLP, TF-IDF, XGBoost, Supervised Learning, Developed a two-stage NLP-based classification model automating mails team's manual process, saving ~$670K annually.
Model Monitoring & Evaluation Framework, Python, Statistical Analysis, KPI Dashboards, Designed monitoring and evaluation metrics for 7 production models (CEOMR, PSI, Legal Text Summarization, OOMF, Protected, etc.).
Retirement Prediction Model, Python, Classification Models, Feature Engineering, Developed a classification model to predict if particular personnel would retire within 12 months.
Hate Protect Model- Hate & Abusive Call Detection, TF-IDF, Logistic Regression, BERT, Automated detection of hate/abusive (HAT) calls previously handled manually.
POC - Generative AI Project Manager, LangChain, LLMs, Prompt Engineering, Python, Built a GenAI-based project manager using LangChain and prompt routing to extract project insights.