Sathwick Reddy Mora

Experience

• 

  Deep Algorithms

  Jan 2021 - Mar 2022

  Data Engineer

  • Configured a data lake with Amazon S3, Redshift, QuickSight, PowerBI, RDS, and Glue for integrated data storage and visualization.• Applied ELT techniques and optimized SQL queries for efficient data retrieval, enabling interactive visualizations with Fusion Charts and PowerBI.• Developed a voice-based Alexa Skill for tracking and managing seizures, boosting patient satisfaction scores.• Implemented Alexa Skill notifications for medication reminders and seizure history checks using sophisticated voice commands and intent functionality.• Integrated Alexa Skill with AWS Lambda using Boto3, streamlining triggers for AWS services. Show less

• 

  Jawaharlal Nehru Technological University, Kakinada

  Jul 2022 - Apr 2023

  Machine Learning Research Assistant

  • Built a GRU based gesture recognition system converting sign language to speech with 95% accuracy, trained on 4 million Word Level American Sign Language video frames.• Implemented memory-efficient Keras sequence techniques, significantly reducing neural network training memory usage by 64%.• Utilized the gTTS library for text-to-speech conversion and OpenCV for image processing while employing MediaPipe for pose estimation, extracting face, pose, and hand landmarks.• Conducted hyperparameter tuning and validation to optimize model performance, prioritizing learning curves and multiple classification metrics to select the best model. Show less

• 

  NarrateAR

  Jan 2024 - May 2024

  AI Software Engineer

  • Architected the backend REST API using the Model-View-Controller (MVC) pattern in C# and ASP.NET for managing lessons, students, and teachers, and integrated the SQL Server database with Entity Framework for data management.• Enhanced engagement for elementary special needs students by creating personal assistants using tailored data for the student and feedback from teachers and parents alongside specialized prompt engineering techniques.• Developed C# interfaces and services integrating OpenAI to create custom lessons with personalized content that assist students in improving their skills.• Crafted a scalable evaluation system utilizing LLMs to assess student progress in lessons, provide detailed performance analytics, and reduce the grading workload for teachers.• Improved LLM response relevance with RAG using OpenAI embeddings and MongoDB vector search to minimize irrelevant responses. Show less

• 

  Buildertrend

  May 2024 - Aug 2024

  Data Scientist

  • Engineered datasets with feature selection from various sources and performed Exploratory Data Analysis to identify peak churn risk factors.• Utilized Google Cloud BigQuery to extract relevant information from various data warehouses using optimized SQL queries.• Recommended a system to track builders' feature usage across various job types, delivering real-time feedback to improve task completion times.• Built an LLM-based system to automate expense tagging, reducing manual reviews and model errors through continuous fine-tuning with a Human-In-The-Loop feature using feedback from the builders.• Minimized manual grouping time by applying clustering techniques, including HDBSCAN and hierarchical clustering on OpenAI embeddings, to identify anomalies and similar expenses across builders' expenses. Show less

• 

  Iowa State University

  Aug 2024 - Dec 2024

  Graduate Teaching Assistant

  • Organized TA office hours for the Programming & Problem Solving through C course, graded assignments, and provided support on foundational programming concepts, helping students build confidence in their coding skills.• Conducted weekly C programming lab sessions, mentoring 25 students to apply coding principles to solve real-world problems, enhancing their hands-on experience and programming skills.• Provided comprehensive support to students in mastering essential concepts such as mathematical modeling, base conversions, Linux commands, and C programming, enhancing their problem-solving capabilities. Show less

• 

  NarrateAR

  Aug 2024 - now

  AI Software Engineer

  • Implemented a multimodal version of lesson interactions, allowing students to submit images along with text to enrich their learning experience and engagement.• Built the backend REST API with C# and ASP.NET, integrating OpenAI LLMs to simplify lesson creation from existing study materials, textbooks, and related documents, all with just a few clicks.• Implemented event logging using Serilog and SEQ to track user flows and analyze interactions, capturing key KPIs such as lesson engagement, token usage, LLM model calls, and cost per lesson.• Developed C# backend controllers to streamline educational workflows for class management, student enrollment, and lesson assignments.• Migrated student conversations from SQL Server to NoSQL database (MongoDB for development, DynamoDB for production), reducing serialization overhead and improving response times.• Crafted a unified C# interface to support seamless integration with NoSQL databases and vector search. Show less