Sep 2024 – Present · 1 yr 9 mos · Pune, Maharashtra, India · On-site

• As a Software Engineer at Equations Work, I specialize in building AI-driven enterprise solutions involving multi-agent orchestration, backend API development, retrieval-augmented generation (RAG), and AI security & evaluation.
• Generative AI & Multi-Agent Systems :
• Designed multi-agent orchestration systems using Azure AI Foundry, LangChain, LangGraph, CrewAI, and Google ADK for intelligent workflow automation.
• Built context-aware chatbots and RAG pipelines with semantic search over multimodal content using PGVector, Google Gemini, and OpenAI.
• Developed AI solutions including synthetic data generation, deepfake detection, and image generation leveraging LLMs and vision models.
• Engineered zero-shot object detection pipelines using OWLv2, Grounding DINO, and YOLOv11.
• Created Model Context Protocol (MCP) servers for real-time data integration.
• AI Security & Evaluation :
• Conducted AI red teaming using PYRIT to identify vulnerabilities and implement guardrails.
• Performed AI agent evaluation using Azure AI Content Safety and Azure Automated Evaluations for production deployments.
• Backend Development & Infrastructure:
• Developed scalable backend APIs using FastAPI, Django, and Flask with DevOps best practices.
• Integrated Azure Cosmos DB, PostgreSQL, PGVector, and vector databases for efficient data handling.
• Built containerized environments using Docker, deployed on AWS EC2 with Terraform.
• Implemented AI orchestration workflows.
• Full-Stack Development:
• Built full-stack applications using React, Material UI, JavaScript, and Python backends.
• Developed interactive AI interfaces with Gradio for rapid prototyping.
• Key Innovations for Experizer.ai:
• Engineered a RAG-powered documentation assistant for Experizer.ai knowledge base.
• Designed an AI content moderation system for Community.Experizer.ai that transforms inappropriate language while preserving context.

Mar 2024 – Sep 2024 · 6 mos · Pune, Maharashtra, India · On-site

• As a Trainee Software Engineer at Equations Work, I contributed to AI-powered applications and GenAI-integrated platforms, working across backend development, LLM integration, and full-stack application design.
• AI & LLM Integration:
• Built multimodal chatbot systems with semantic search over video content using RAG pipelines and LangChain.
• Developed Flask and FastAPI APIs integrating LLMs (ChatGPT, Anthropic, Google Gemini, Stable Diffusion) with PostgreSQL for diverse applications including data augmentation and content generation. - Created AI-powered content moderation systems using Google Gemini for intelligent text transformation.
• Designed RAG-based pipelines for video content processing enabling context-aware information retrieval.
• Backend Development & Database Design:
• Engineered backend APIs using Flask, FastAPI, and Django for scalable AI-driven applications.
• Designed and optimized PostgreSQL and MongoDB databases to meet client-specific requirements.
• Implemented SSO authentication systems bridging MySQL and MongoDB architectures for seamless user management.
• Synthetic Data Generation Platform :
• Led the full development lifecycle of a synthetic data generation product using few-shot prompting with LLMs for data augmentation and privacy-preserving machine learning.
• Managed end-to-end product development from Figma mockups to production deployment.
• Built scalable backend infrastructure supporting diverse data generation workflows.
• Full-Stack Development:
• Designed and developed user interfaces using React.js and Material UI, enhancing user experience across multiple projects.
• Built responsive, intuitive frontend applications integrated with Python-based backend services.
• Contributed to performance optimization and feature enhancements across distributed systems.
• Through these efforts, I gained hands-on experience in AI-driven development, GenAI integration, RAG architectures, full-stack solutions, and scalable software design.