Shamli
Ingole

Software Engineer 2

● Spearheaded backend development from infrastructure to API layer by provisioning DynamoDB via CDK and implementing GraphQL and Spring Boot endpoints to support reliable CRUD operations for configurable reporting workflows. ● Wrote Java unit tests using Mockito for multiple Spring Boot services, reaching 100% coverage and improving overall reliability.

Software Engineer - AI/ML

● Engineered a PySpark batch pipeline using Azure OpenAI’s Global Batch API to process large-scale multimodal product data (images + text). Cut LLM inference costs by 50%, and enriched HANA metadata , boosting product recommendation accuracy. ● Enabled multi-image support to improve extraction depth and leveraged prompt engineering to combine textual descriptions with images, significantly boosting classification accuracy. ● Integrated LangChain with HANA for semantic vector search, improving attribute extraction accuracy in Product DNA analysis from 68% to 85%, and enhancing catalog metadata enrichment and product clustering. ● Developed and optimized comprehensive unit tests, ensuring robust coverage across all vision service use cases and enhancing system reliability. ● Built a Visual Search feature using OWL-V2 for object detection, enabling object-level image-based product discovery and improving search accuracy across catalog inventory. ● Built a configurable Agentic AI shopping assistant using FastAPI Python and LangChain, enabling dynamic persona-driven conversations, secure prompt handling, and vector-based product recommendations that accelerated product discovery by 3x . ● Developed a Redis-backed asynchronous job queue to process 10k+ product attribute extraction requests, leveraging PostgreSQL to track job status and retries. Reduced processing time and eliminated task failures at peak load by enabling parallel worker scaling.

Software Engineer

● Built a .NET (C#) desktop application using OOP and design patterns to automate config backups from MongoDB and MySQL databases across 7 stages. Replaced 32 manual steps and brought backup failure rate to zero. ● Developed KDBot, an LLM powered enterprise knowledge retrieval system using RAG, incorporating domain-specific fine-tuning for improved relevance. Boosted customer interaction efficiency by 40% through enhanced natural language response quality.

Technology Specialist

Software Engineer

● Designed Python Flask APIs to automate setup of 36+ nested Google BigQuery config tables and trigger Google Cloud Composer DAGs, eliminating 3+ hours of manual setup per machine learning pipeline across 20+ workflows. ● Developed Python FastAPI microservices for Google’s Document Intake Accelerator to automate classification and extraction for 100K+ documents, improving throughput from 3 to 15 batches/hour and reducing human review time by 60%. ● Implemented CI/CD pipelines with Skaffold, Docker, and GitHub Actions for containerized microservices on Google Kubernetes Engine (GKE). Reduced deployment time from 25 to under 10 minutes, improving release speed and reliability. ● Built an Angular-TypeScript web app for Southern New Hampshire University, enhancing student engagement by 70% among 2 million users. Ensured seamless scalability and achieved 95% user satisfaction through rigorous UI testing and component optimization. ● Designed and incorporated React components, seamlessly integrating Redux for efficient state management. This optimization resulted in a substantial 47% reduction in application load time. ● Orchestrated a scalable microservice for a pharmaceutical client, leveraging machine learning (85% classification, 70% extraction accuracy) and integrating Google Firebase and Cloud Task, significantly enhancing regulatory information retrieval efficiency. ● Increased test coverage from 45% to 90% using Pytest for unit and end-to-end tests, catching critical bugs pre-release. ● Awarded as trailblazer for exceptional contribution to the development of a project as a software developer, demonstrating strong leadership skills and the ability to innovate and drive progress.

Software Engineer Full Stack

● Led backend development of a high-traffic e-commerce platform using Django REST, AWS Lambda, Redis, and Elasticsearch, scaling to 100K+ monthly users and reducing API latency from 600ms to 350ms by optimizing caching and search performance. ● Built a Python test automation framework for data pipelines, enabling 20+ QAs to eliminate 6+ manual validation steps per run. ● Implemented Git for version control, Jest for frontend tests, and Swagger for API documentation. Reduced conflicts by 30%, increased test coverage by 40%, and enhanced documentation clarity by 20%. Resulting in refined code management. ● Led an Agile team in developing a sophisticated web application utilizing Django REST framework, React (JavaScript Library), HTML, and CSS, enabling the creation of essential features for a top-tier, user-centric application. ● Awarded top performer in Python training batch, demonstrating exceptional understanding and practical application of the language. Contributed to web apps and data analysis tools, adding value to the software development team.

Software Engineer

● Built a web application using JavaScript, HTML, and SCSS, creating a feature-rich blogging platform. Users can author their blogs and engage with content from others, including liking and commenting, but only after logging in. The site allows blogs to be categorized for easier navigation. ● Developed a 3-tier web application "Shout-It," utilizing React.js for the front end and Python with Django REST framework for back-end development. This RESTful service enables users to express themselves through shouts in various formats—text, audio, video, image. Features include the ability to update and delete shouts, send friend requests, and view shouts from friends.

Software Engineer

● Built a Spring Boot microservice in Java to track real-time data from 100+ machines, improving fault detection speed by 3×. ● Designed a Kafka pipeline to process 500K+ events, enabling real-time alerts and cutting equipment issue response time by 70%. ● Authored comprehensive unit tests using JUnit for ensuring robust code quality and functionality. Enabled real-time manufacturing process and equipment monitoring in a factory setting.