Hi, I'm |

Key Achievements

• Haleon GenAI Assistant: Architected and developed an enterprise-grade, multi-team GenAI platform integrating FastAPI, React, Azure OpenAI and Azure Cloud Services to enable secure, intelligent data access across Finance, Procurement, and Help Center from scratch.

  • Built team-specific AI agents and modular tool frameworks using LangChain and LangGraph for contextual data retrieval.
  • Enhanced retrieval quality through semantic ranker with multilingual support and hybrid search using text and vector search in Azure AI Search.
  • Engineered a scalable, structure-aware document processing pipeline using Azure AI Document Intelligence with automated GitHub Actions workflows, extracting structured layouts and entities, performing chunking, ingesting and managing documents in Azure Blob Storage, and updating vector and keyword indexes in Azure AI Search for real-time, high-relevance retrieval and RAG-ready content delivery.
  • Experimented with MCP (Model Context Protocol) client–server architecture for distributed tool orchestration across databases and AI services.
  • Optimized chat architecture for sub-second latency through asynchronous processing, containerized deployment, and scalable microservice design.
• Prometheus: Implemented end-to-end MLOps framework using Databricks Bundles, Unity Catalog, and GitHub Actions. Built robust CI/CD workflows with linting, bundle validation, pre and post semantic versioning, and automated Databricks job orchestration across Dev/UAT/Prod. Enabled dynamic, market and brand specific task generation at runtime, versioned model and data persistence in Unity Catalog, reproducible deployments via explicit semantic release tags, with outputs seamlessly refreshed in Power BI for stakeholders.
• Federated Help Center Search: Architected and delivered a federated search solution across 15+ brand help center URLs, providing search results from multiple brand websites and help centers across different markets, languages, and brands.

  • Large-scale web crawling from 500+ brand websites (including help centers) across multiple countries, markets, domains, and languages. Automated scraping using Scrapy, structure-aware information extraction from leaflets using Azure AI Document Intelligence, data ingestion to Azure Blob Storage, and updating vector and keyword indexes in Azure AI Search for real-time, high-relevance retrieval and RAG-ready content delivery.
  • Providing robust search, autocomplete, and suggestion APIs from Azure AI Search directly to the frontend team for seamless integration.
  • For internal testing using GenAI assistant: Added Playground mode with custom model selection, temperature control, and system prompt configuration, including full prompt visibility alongside search results for enhanced experimentation and debugging.

  Federated Search API Demo

  Interactive demo of Autocomplete, Suggestions, and Search APIs Integrated across 15+ help centers providing results from brand websites

  Federated Search Demo

  Scroll into view to see animation
• Procurement Contract Analysis: Led ML-Ops for a RAG-based (Agentic-RAG) system by productionizing modular workflows with Databricks Asset Bundles, integrating Azure Document Intelligence, Unity Catalog, and model serving endpoints along with automated deployment of Databricks App. Built CI/CD pipelines with GitHub Actions for automated, rollback-ready workflows, deployments, and GitOps-based orchestration.
• Asset Vision: Designed, implemented a robust CI/CD pipeline with GitHub Actions to build and publish Docker images to GitHub Container Registry (GHCR) and Azure Container Registry (ACR), enabling automated deployment to Azure Web App and microservices on Kubernetes cluster for scalable, secure and reliable application management across environments. Applied same approach to streamline deployments in other RAG based project.
• MLOps Template for Databricks: Improved and implemented a comprehensive MLOps template for Databricks leveraging GitHub Actions to automate CI/CD pipelines, including code linting, testing, environment setup, and seamless deployment of Spark jobs, such as model training, validation, and deployment, resulting in significantly enhanced workflow efficiency and reduced deployment times.

🏆Awards & Recognition

• Global Employee Recognition

  BronzeUnlock Value at PaceDec 2025

  Rajesh has 1168 GitHub contributions this year, which is impeccable. He led MLE efforts and delivered successful deployments for Help Centre Federated Search, Prometheus, and Contract Analyzer projects. He presented the Contract Analyzer project on AI Day, receiving significant recognition across the Data Office and CDO.

• Global Employee Recognition

  BronzeGo BeyondNov 2024

  Outstanding ML Engineer with major contributions to GenAI initiatives including Help Centre Federated Search and RAG-based chatbots using Azure and Databricks. Significant impact on MLOps through Prometheus and MLOps template contributions. Recognized as a strong team player, actively supporting deployments and leading discussions across MLOps forums.

• Global Employee Recognition

  Collaborate for ImpactSep 2025

  Exceptional contribution and energy during Career Day and Technovation Day. Presented Contract Analyzer and Help Centre Federated Search projects, receiving widespread appreciation across the Data Office and CDO.

• 2025 Excellence Award – India Capability Centre

  Simplify for ImpactWin as OneDec 2025

  Helped bring Haleon’s “Win as One” strategy to life. 2025 Excellence Award – India Capability Centre for Build Contract AI – A synergy of Procurement Analytics and Data Office, revolutionising contract intelligence and management using Agentic-RAG.

⚙️Technologies Used

Key Achievements

• CI/CD Infrastructure: Robust CI/CD pipelines using Cloud Build for seamless and automated deployment of microservices and batch services on GKE, Cloud Functions, and Cloud Run, significantly reducing deployment errors, manual effort, and turnaround times across environments. Collaborated with DS team to optimise code and model inference for higher throughput.
• EdgeDB Inference Store: Architected, deployed and upgraded Inference store (EdgeDB on Google Kubernetes Engine with provisioned PostgreSQL instance as backend) with FastAPI endpoints deployed in GKE. API employed across various DS Services for efficient caching and retrieval of inferences in real time (latency around 5ms), as well as in batch processing jobs with traffic coming from Queue (PubSub) or an Orchestrator (Airflow). Integrated CI/CD pipelines for automated deployments in GCP (Upgrade Flow).
• Feast Feature Store: Efficiently deployed Feast feature-server on Google Kubernetes Engine, leveraging an automated materialization pipeline for real-time data updates in Redis (Online Store) and BigQuery (Offline Store) via Cloud Functions triggered by a diverse range of team data uploads in a GCS bucket. API effectively addresses the challenge of accessing up-to-date features from CSV files and currently employed in 10+ DS services, delivering features with minimal latency (<10ms). CI/CD enabled for robust rollouts.
• Data Flow Pipeline: Built data flow pipeline to optimize the timely data sync between Snowflake and Google Sheets using Airflow, Pub-Sub, EdgeDB and GKE. Syncing data for approx 2500 Sheets with roughly 15000 Jobs/day. (blog) (citation by Modern Data Stack, citation by Data Engineering Weekly).
• Airflow Orchestration Platform: Deployed, administered, upgraded a robust Airflow infrastructure on Google Kubernetes Engine, overseeing 70+ diverse git synchronised DAGs running approx 1000+ times daily, covering a total of 3800+ tasks executions to support various DS-BI teams. Efficiently managed the execution of tasks, encompassing Cloud Function triggering, job scheduling, PubSub messaging, EdgeDB integration for jobs status update, and Snowflake-to-Google Sheets data synchronization and seamless integration with diverse GCP services. Implemented RBAC to enhance team visibility, established automated log cleanup with ensuring a streamlined, well-maintained system.
• NBFC Infrastructure Migration: Spearheaded migration of NBFC DS infrastructure from AWS to GCP, breaking down monolith into scalable sub-monoliths and GKE-based microservices. Decoupled DB operations, slashing response times from 3 seconds to 503ms (99th percentile), eliminated overall timeouts, previously at 8-9%. Implemented streamlined data flow via Pub/Sub and Cloud Functions, enabling efficient EdgeDB updates and hourly Snowflake synchronization. CI/CD pipelines orchestrated for seamless deployment of microservices on GKE and respective cloud functions.
• Triton Python SDK: Developed a pip-installable, reusable Python wheel package for seamless integration with Triton Inference Server, supporting high-performance model inference by handling both HTTP and GRPC requests in synchronous and asynchronous modes, with built-in logging, error handling, and configuration flexibility.
• Request-Response Store: Designed, deployed high-performance APIs Request-Response Store using FastAPI, featuring client authentication, rate limiting and low 5ms latency. Orchestrated storage of request-response in EdgeDB, synchronized to Snowflake based on team-defined frequencies, and leveraged by 10+ DS services.
• Customer Interaction Analytics: Designed and implemented a cloud-based workflow for end-to-end processing of customer interactions data. The workflow encompasses data ingestion from Snowflake, audio download, transcription generation, information extraction, and automated customer summary generation. This system is orchestrated through Airflow and leverages Google Cloud Functions, Cloud Storage, Kserve InferenceService, Triton, MongoDB and EdgeDB.
• Data Platform R&D: Experimenting with deploying Doris, StarRocks, and associated open-source tools on GKE to construct a scalable and high-performance data platform, incorporating S3-compatible object storage via MINIO/GCS, implementing access control and data retention policies, and ensuring effective Kubernetes cluster monitoring and alerting through Grafana's dashboards utilizing Prometheus as a primary time-series data source.
• Audio Embeddings Search: Experimented Minio and Milvus integration to extract audio embeddings from Minio, enabling efficient search for similar content, Redis for metadata storage.

🏆Awards & Recognition

• Rookie Award – Appreciation Certificate

  Special RecognitionQ3 2022

  This certificate is proudly awarded to Rajesh Dhanda (ML Engg – Global) for outstanding performance in Quarter 3, 2022.

  Signatories: VP – DS and BI, COO

⚙️Technologies Used

Blogs Written & Contributed

Key Achievements

• Implemented a reference direction-based entry-exit system and Object Counting Feature using DeepSort, showcased in video.
• Conducted TensorFlow Lite model conversion of DeepSort and executed its deployment on Raspberry Pi hardware.

⚙️Technologies Used

Video Showcase

Directions & Counting of moving Vehicles with velocity vector using Deep SORT

Click to watch on YouTube

Key Achievements

• Developed algorithm for many-to-many mapping of symptoms to diseases, enabling disease diagnosis based on symptom sets.
• Created mathematical formulation to facilitate symptom recommendation, improving user experience and healthcare decision-making.
• Generated random datasets to rigorously test and optimize ML algorithms for recommendation systems.

⚙️Technologies Used