Agentic AI & Orchestration | Full-Stack Backend Architecture | MLOps & Production Engineering
Senior AI & Systems Engineer with 4+ years of designing and deploying scalable Agentic AI, RAG pipelines, and cloud-native backend architectures. Proven track record leading cross-functional AI initiatives, optimizing autonomous workflows with LangGraph, and deploying secure LLM microservices across AWS and Azure.
| 🤖 Agentic AI & Orchestration | Multi-Agent Orchestration, Autonomous Planning & Execution Loops, Proactive Memory & Tool-Use Systems, LangGraph, LangChain, DSPy, Advanced Prompt Optimization |
| 🧠 ML, Forecasting & RAG | Contextual AI, Retrieval-Augmented Generation (RAG), Vector & Similarity Search (Pinecone, Weaviate), Multi-Model Ensembles, CNN Image Classification, PyTorch, TensorFlow |
| ☁️ Cloud & MLOps | AWS (Bedrock, SageMaker, EC2, S3), Azure, GCP, Containerization (Docker, Kubernetes), CI/CD (Jenkins, GitLab), Real-Time Performance Monitoring, System Reliability |
| 💻 Core & Backend | Python (Asyncio, Pydantic, FastAPI), System Architecture, Software Development, APIs, Database Design (SQL, PostgreSQL), Knowledge Graphs (Neo4j, Graphiti) |
Dentsu Global Services | Mumbai, India (May 2022 – Present)
- Architected Hierarchical Multi-Agent Systems: Deployed an autonomous agent ecosystem via LangGraph with 3 specialized agents (Execution, Review, and Feedback loops), reducing production cycle times by 86%.
- Enterprise GenAI & Document Extraction: Developed and integrated AI/ML models, APIs, and document extraction on AWS Bedrock, mitigating manual data entry workflows by 52%.
- Technical Leadership & Standards: Mentored AI engineers, established development strategies, conducted code reviews, and enforced rigorous coding & quality benchmarks.
- Reliability & Performance Tuning: Monitored system performance, security, and scalability, utilizing latency tuning and Root Cause Analysis (RCA) to eliminate pipeline bottlenecks.
- Full-Stack GenAI Pipelines: Designed and scaled production GenAI workflows leveraging GPT and Gemini models to automate contextual enterprise processes.
- Cloud ETL Pipelines: Engineered robust, high-throughput ETL data pipelines to ingest, clean, and structure large-scale datasets across cloud warehouses.
- R&D Leadership: Evaluated and integrated cutting-edge AI developer tools, leading R&D on GitHub Copilot strategies adopted globally.
- Cluster Data Ingestion: Built containerized Python web scraping and data crawling pipelines optimized for cluster-based processing, streamlining massive e-commerce ingestion.
- Enterprise API Integration: Integrated custom data pipelines seamlessly with enterprise APIs during rapid-prototyping Agile sprints.
| Project | Highlights & Architecture | Impact Metrics |
|---|---|---|
| Hierarchical Multi-Agent Ecosystem | LangGraph orchestration combining Execution, Reviewer, and Feedback loops with DSPy programmatic prompt optimization. | 🚀 86% Cycle Time Reduction |
| AWS Bedrock Intelligent Document Extraction | Serverless LLM pipeline on AWS Bedrock automating unstructured document comprehension and structured data extraction. | ⚡ 52% Manual Workflow Mitigation |
| Enterprise Production LLM RAG Pipeline | Low-latency, context-aware RAG pipeline utilizing advanced hybrid vector similarity search with LangChain & vector databases. | 🎯 High-Precision Context Retrieval |
| ML Model Microservice Deployment | Trained CNN model in TensorFlow, containerized and deployed as a low-latency Dockerized REST API for real-time validation. | ⏱️ Sub-second Inference Latency |
| End-to-End Business Automation | Full-stack automated business process workflows combining custom Python scripts and enterprise connectors. | 📉 40% Operational Effort Reduction |
Auto-curated and synchronized via GitHub Actions cron based on repository activity and topics.
| Project | Description | Tech Stack | Stats | Updated |
|---|---|---|---|---|
| Drishti 🌐 Live Demo |
🛡️ Real-time Agentic Scam & Fraud Intelligence for Banks - Intercepting APP fraud (Digi... | |
⭐ 0 🍴 0 |
Oct 18, 2026 |
| Kaleido 🌐 Live Demo |
One source, a verified experience for every audience. Multi-format personalization & fa... |  |
⭐ 0 🍴 0 |
Oct 15, 2026 |
| JanSetu | JanSetu is a sovereign Digital Public Good (DPG) that turns citizen grievances across B... | |
⭐ 0 🍴 0 |
Sep 30, 2026 |
| tanglefoot | A rigorous public benchmark and interactive dashboard evaluating LLM-based agent framew... | |
⭐ 0 🍴 1 |
May 31, 2026 |
| mintwright | The self-evolving agent that builds its own tools at runtime. | |
⭐ 0 🍴 0 |
May 17, 2026 |
| portfolio | Autonomous AI & Generative AI project. | |
⭐ 0 🍴 0 |
Apr 20, 2026 |
- 🔨 Pushed to Drishti:
Pushed new updates— Oct 16, 2026 - 🔨 Pushed to JanSetu:
Pushed new updates— Sep 30, 2026
|
📜 Professional Certifications & Accreditations • AWS Certified: Machine Learning – Specialty (Amazon Web Services) • Microsoft Certified: Azure AI Fundamentals (Microsoft) • Google Cloud: AI Agent & ADK Builder Accreditations (Explore Google Cloud) • Data Science & Deep Learning Professional Certificates (IBM / Credly) |
🎓 Education • Bachelor of Engineering in Information Technology (2018 – 2022) M.H. Saboo Siddik College of Engineering, Mumbai, India CGPA: 8.48 / 10.00 |
⚡ *Automated with GitHub Actions • Last synced: `2026-10-18 03:57 UTC`*