I lead the development of AI-driven agent architectures and full-stack software systems, integrating generative AI into our tech stack. My work includes engineering microservices, RESTful APIs, and front-end interfaces to ensure seamless interaction between users and AI agents. By deploying these AI Driven Software solutions, I’ve significantly enhanced system efficiency, user experience, and software quality. My focus is on improving efficiency, speed by reducing manual intervention through Generative AI and Agentic Workflows using LangGraph

Key Highlights:

Database Interaction: Developed agents that interact with SQL/NoSQL databases, constructing SQL queries from natural language to perform data operations seamlessly across back-end systems.

GitHub Workflows: Created AI agents that perform PR reviews and automate CI/CD processes, integrating with microservices and API gateways to streamline code analysis, deployment, and registry updates, reducing deployment times by 40% and improving overall code quality.

Code Analysis: Developed agents that generate temporal and predicate logic for spec design and continuously analyze code coverage using control flow graphs. By leveraging previous project data, these agents have reduced the time required for new project setup, improved code quality and saved significant time and resources.

Chatbot Integration: Built a web based chatbot that connects users with the AI agents, allowing for real-time interaction with tools, databases, and the requirement management system, enhances accessibility by providing users with immediate access to relevant company wide information, significantly reducing manual search efforts.

Advanced AI Integration: Deployed AI agents using LangChain, Crew AI, and AutoGen, optimized with fine-tuned LLMs like Mistral-NeMo and Llama 3.1. This continuous operation across distributed systems has improved overall system performance, saved costs, & contributed to faster, more reliable project deliveries.

I've architected and developed a versatile suite of AI-driven applications tailored to enhance internal operations and decision-making across various domains. These applications leverage advanced NLP and fine tuned LLM's to provide contextual and intelligent support, thereby enhancing productivity and information accessibility:

InSite Hub: A RAG-based semantic search tool that provides a unified search interface across multiple internal repositories, including Confluence, SharePoint, and an extensive library of documents. This tool helps employees access and retrieve organizational knowledge, significantly reducing time spent on information searches and enhancing productivity across the board.

Meeting Assistant: This tool is seamlessly integrated with the Microsoft Teams API, designed to manage meeting transcripts efficiently. It automates the retrieval of transcripts and enables detailed Q&A about meeting content, The assistant provides comprehensive summaries and actionable insights, transforming how meetings are conducted and reviewed, it enhances meeting effectiveness.

Universal Assistant: A general-purpose chatbot designed to address a broad range of queries, significantly boosting workplace productivity by providing quick and accurate answers to everyday questions.

JIRA Assistant: Designed a robust Q&A system for organizational JIRA queries, incorporating semantic processing to understand user queries related to the entire JIRA setup. The assistant processes the JIRA warehouse schema, forms SQL queries, and directly interfaces with JIRA APIs to retrieve and present real-time data for users.

Leadership & Collaboration:

Global Collaboration: Coordinated with international teams, aligning development processes and project outcomes across different time zones and cultures.

Stakeholder Communication: Engaged regularly with stakeholders to align projects with business objectives &ensure satisfaction.

Key Project Implementations:

DataLens: Was Part of the development of a Business Intelligence tool that improves data accessibility and decision-making through dynamic data visualization and analytics.

STAR (System Test Automation Repository): Created an automation system to streamline firmware validation, connecting devices running test cases & reducing manual testing.

FAR (Failure Analysis and Reporting): Developed a system that automates data extraction and analysis from firmware test logs, enhancing failure analysis speed.

Contributions:

Frontend Development: Developed BI dashboards with interactive components using Angular, enabling intuitive navigation and real-time data-driven decision-making.

Backend Development: Built robust micro-services with Python and .NET Core, supporting data-intensive applications with REST APIs & GraphQL, deployed in Docker containers managed with Kubernetes for scalability.

Database Management: Developed data warehouses, wrote complex stored procedures, enhancing data retrieval speeds and supporting high-performance data processing.

Scalability & Reliability: Implemented dynamic scaling with KEDA and advanced data management strategies to handle increasing volumes efficiently.

Distributed System Development: Led the design and deployment of a global distributed system, employing advanced caching and data replication to ensure high availability and optimal performance.

System Monitoring & Performance Optimization: Utilized Elasticsearch and Prometheus for real-time monitoring and performance metrics visualization.