Sahithi Pendry Email & Phone Number

San Francisco, Ca, Us

San Francisco Bay Area

Buffalo, New York, United States

San Francisco Bay Area

- Architected and developed 360-degree customer and user metadata models using DBT (Data Build Tool), consolidating key insights from Salesforce and Netsuite, enabling more actionable data for stakeholders.- Automated data pipelines using Fivetran and Databricks to extract and transform data from Lessonly, reducing manual reporting processes by 3 hours weekly and improving customer experience through real-time insights.- Designed scalable ERD diagrams and normalized database schemas based on user requirements, optimizing data structures for better performance and accessibility.- Collaborated with business teams to translate operational requirements into data models, facilitating efficient data flow across cloud-based environments.- Employed SQL and PySpark for complex data transformations and aggregation, ensuring the accuracy and consistency of integrated datasets.- Enhanced ETL (Extract, Transform, Load) processes and improved data governance practices, ensuring compliance with organizational data standards.

London Area, United Kingdom

As a Machine Learning Engineer at Forensic Alpha, I built microservices that calculate red flags hidden deep within financial statements and extract, clean, format, validate, and store raw data in MongoDB to generate machine-generated reports using NLP. I also executed a generic Python framework to extract various data items from Annual Reports and implemented a framework on top of XBRL formatted Annual Reports of the US Region to fetch data points. I researched models like BERT, RoBERTa, DistilBERT, and XLM-RoBERTa to devise a Long Context Question Answering system that retrieves data from paragraphs in Annual Reports across Europe Region. Additionally, I accomplished an event-based architecture where all the microservices are connected via step functions in AWS, and the codebase is connected to Lambda (AWS service) via Git.

Chennai Area, India

As a Machine Learning Engineer at Freshworks, I conducted research on feature engineering and built machine learning models using various embeddings such as LASER, Multi-Lingual Universal Sentence Encoder, Universal Sentence Encoder, and Sentence Transformers. I also created a script to upgrade TensorFlow from version 1 to version 2 in the official Google BERT Code to fine-tune Sentence Transformers Model. Additionally, I fine-tuned Hugging Face Models to automatically detect the thank-you intent in tickets and close them. I achieved a logistic regression model to remove email greetings and footers so that the body can be used as a FAQ. I also developed a chatbot integrated into Microsoft Teams to help solve IT queries and tickets raised by employers. Finally, I designed models to help agents reply faster than usual in the chatbot, including Smart Compose built on Elastic search to auto-complete the sentence, and Smart Chat by the TensorFlow model that provides the top three suggestions to reply to.

Chennai, Tamil Nadu, India

As a Product Development intern, I designed and implemented five integrations for various billing and invoicing systems, including Zuora and Chargify. In addition, I created a custom invoice generator for a client that needed a specialized solution. My work was recognized by Sotheby's, who rewarded me with a gift hamper for completing the project within three days.

Gujarat, India

During my time at TopHawks, India's premier sales outsourcing, visual merchandising, mystery shopping, and BTL promotions company, I worked on revamping the company's website. This involved redesigning website pages, curating content and creating custom images to meet specific requirements.

India

Thanjavur, Tamil Nadu, India

Master'S Degree, Computer Science, 3.66

B-Tech, Ict, 8.5