SangeethaYemisetty
A Passionate Machine Learning Developer with nearly 4 years of experience in building ML models for various industries targeting healthcare, banking and ecommerce. Expertise in Time Series analysis, clustering & classification, web scraping, data analysis and visualization. Started as an Individual contributor, Currently leading a team of 4 members and is a key architect in developing Machine Learning solutions. Flexible and versatile to adapt to new environments and technologies. Strong communication, interpersonal, learning and organizing skills matched with the ability to manage the stress and time effectively.
Skills
- Design and develop highly scalable Machine learning models.
- Consulting and troubleshooting support for all the adoption teams in the department.
- Monitoring model performance and hypertuning to improve the relevancy and compatibility.
- Participation in Code reviews and Design discussions.
- Developing utilities and generic reusable solutions used across applications
- Platform Integration testing & Debugging
- Review the code and ensure proper best practices are followed across the products
- Mentoring the team members.
- Agile Development & Scrum
Awards
- Best performer of the team – Q4’20.
- Won people's choice award. –Resource Management using AI.
- Best performer of the team – Q2’19.
- STAR Award – For driving the EPIC effectively.
PROFESSIONAL EXPERIENCE
Brane Enterprises, Hyderabad – Sr. Machine Learning Developer
Nov 2020 - PRESENT
Pegasystems, Hyderabad – Machine Learning Developer
July 2017 – June 2020
Project profile
Brane Enterprises, Hyderabad - Sr. Machine Learning Developer
Skills Used: NLP, Web scraping, ARIMAX, SARIMAX, Infersent, Python, Django, Faiss indexing, FAST API, Matplotlib, Seaborn, PCA.
Projects: Predictive Analytics, Measure Generation, Analytics Engine.
Roles and Responsibilities:
- Developed the base architecture combining both generated and library measures.
- Analysed ARIMAX, SARIMAX models for recommending relevant business analytics.
- Improved the measure recommendation relevancy using NLP techniques.
- Complete integration of ML models using FAST API and Django.
- Automated the ETL(extract, transform, load) process.
- Curated thousands of records of measures using Web scraping.
- Used FAISS indexing to improve the performance of Nearest neighbours search.
- Automation script for testing the flow with bulk no. of use cases.
- Using Matplotlib, Seaborn developed the Analytical dashboard.
- Implemented the caching mechanism for performance optimization.
- Collaborated with various stakeholders in integrating the Analytical engine with platform.
- Lead & mentored a team of 4 members in improving the code quality and reusability.
- Single point of contact for the entire Analytical engine.
- Review the code and ensure proper best practices are followed across the products.
Pegasystems, Hyderabad - Machine Learning Developer
Skills Used: Recurring Neural networks, NLP, LSTM , Support Vector Machines, Random Forest, Logistic regression, ARIMA, Market Basket Analysis.
Projects: Chatbot for a banking company, Loan approvals estimation, Sales Prediction.
Roles and Responsibilities:
- Carried out data extraction and data manipulation using Python, Spark ,SQL tools.
- Handled the word embedding mechanisms like tokenizing, stemming, and lemmatization.
- Built the Seq2Seq Neural network using Tensor flow.
- Applied boosting algorithms like bi-directional LSTMs, bucketing to improve the bot performance.
- Predicted the Default and Non-default customers in the next 3 months with 95% accuracy using the ML algorithms like Logistic regression, Decision trees and Random forest.
- Validated the test data using techniques like AUC, Sensitivity, Specificity, F1-Score, Decile analysis.
- Applied boosting algorithms like Adaboost, XGBoost to improve the performance.
- Market Basket Analysis: Used the partner data to develop market baskets predicting the likelihood of products being purchased together and customize offers to cross sell the products .
- Data Visualization using MATPLOTLIB, and ARIMA time series model for forecasting the sales
Education
BITS, Pilani
Vasavi College of Engineering
Interests
Sports
Browsing Internet
Listening Music
Gadgets ❤️
Travelling