Rahul V.

Data Engineer and Analytics Engineer building ETL and ELT pipelines, cloud data warehouses, governed BI and production AI, handed over as code your team owns. Most data work fails at the seams. The pipeline lands data nobody trusts. The dashboard shows a number finance disputes. The AI pilot stalls because the data model underneath it was never built. I cover data engineering, business intelligence and machine learning end to end, so those seams stay closed. 15 delivered projects across retail, e-commerce, finance, oil & gas and manufacturing. Founded 2021. Team of 10. 𝗗𝗔𝗧𝗔 𝗘𝗡𝗚𝗜𝗡𝗘𝗘𝗥𝗜𝗡𝗚: 𝗘𝗧𝗟, 𝗘𝗟𝗧, 𝗣𝗜𝗣𝗘𝗟𝗜𝗡𝗘𝗦, 𝗪𝗔𝗥𝗘𝗛𝗢𝗨𝗦𝗘 𝗠𝗜𝗚𝗥𝗔𝗧𝗜𝗢𝗡 Cloud data pipelines, lakehouses and data warehouses on Azure, AWS and GCP: Azure Databricks, Snowflake, Google BigQuery, Amazon Redshift, Azure Synapse, Azure Data Factory, AWS Glue, Apache Airflow, dbt, Kafka, Flink, Spark and PySpark. New builds, plus data migration off SSIS, Informatica, Oracle and end-of-life on-premise estates, with parallel-run reconciliation so you can prove the new warehouse matches the old before cutover. Real-time streaming pipelines, change data capture and dimensional data modelling that stays queryable as volumes grow. 𝗕𝗨𝗦𝗜𝗡𝗘𝗦𝗦 𝗜𝗡𝗧𝗘𝗟𝗟𝗜𝗚𝗘𝗡𝗖𝗘: 𝗣𝗢𝗪𝗘𝗥 𝗕𝗜, 𝗧𝗔𝗕𝗟𝗘𝗔𝗨, 𝗟𝗢𝗢𝗞𝗘𝗥 Power BI dashboards, Tableau reporting and Looker semantic models on one governed metrics layer, so teams stop reporting three different revenue figures. Data governance with named metric owners, dbt test coverage, data quality checks and data contracts in CI, plus observability that catches incidents before your users do. Embedded, multi-tenant analytics for SaaS products. 𝗔𝗜 𝗔𝗡𝗗 𝗠𝗔𝗖𝗛𝗜𝗡𝗘 𝗟𝗘𝗔𝗥𝗡𝗜𝗡𝗚: 𝗥𝗔𝗚, 𝗟𝗟𝗠 𝗔𝗚𝗘𝗡𝗧𝗦, 𝗣𝗥𝗢𝗗𝗨𝗖𝗧𝗜𝗢𝗡 𝗠𝗟 GenAI and LLM agents that cite their sources and escalate instead of guessing. RAG pipelines, vector databases and agent orchestration with LangChain, LangGraph, CrewAI, Qdrant, Milvus, Whisper and FastAPI. Plus production machine learning: demand forecasting, feature platforms and MLflow-managed retraining, with guardrails, labelled evaluation and a regression gate before release. 𝗦𝗘𝗟𝗘𝗖𝗧𝗘𝗗 𝗥𝗘𝗦𝗨𝗟𝗧𝗦 • Retail data warehouse migration: 480+ SSIS packages on SQL Server rebuilt as an Azure Databricks lakehouse. Nightly batch 7h40m to 1h05m, SLA breaches 11 nights a month to under 1. • Oracle Exadata and Informatica retired onto Azure Synapse for an insurance carrier. Nightly ETL load 9h20m to 1h50m, recovery from a full 9-hour rerun to restarting at the failure point. • Regulatory reporting data mart for a bank: reporting cycle 11 working days to 2.5. • Real-time inventory and store operations platform: stock latency from overnight batch to under 5 seconds. • Embedded customer analytics for a B2B SaaS product: in-app dashboards 20s+ at P95 to under 3 seconds, tenant isolation structural and CI-tested. • Enterprise Power BI rollout across a manufacturer's plants: month-end prep from 3 days per plant controller to same-day. • Production and HSE analytics for an oil & gas operator: daily operations pack from manually assembled at 10:00 to automated by 06:30. • AI support agent for a B2B SaaS platform: 58% of qualified tier-1 intents resolved autonomously, with citations and escalation. • Analytics assistant and demand forecasting for a retailer: ad-hoc data team requests from ~340 a month to ~95. 𝗛𝗢𝗪 𝗜 𝗪𝗢𝗥𝗞 You own the code: version-controlled, tested and documented, in your repositories and cloud accounts. No proprietary framework to keep paying for. I lead with the business problem, not the tool. If three teams report three different revenue numbers, the fix is a governed definition and an owner, not another dashboard. Quality gates block bad data rather than publishing it quietly: data contracts in CI, freshness checks, tests that fail the pipeline not the month-end close. Handover is part of delivery: runbooks, architecture docs and enablement, so your team ships without me. 𝗧𝗘𝗖𝗛 𝗦𝗧𝗔𝗖𝗞 Azure Databricks, Snowflake, BigQuery, Redshift, Synapse, Delta Lake, Unity Catalog, dbt, Airflow, Kafka, Flink, Spark, PySpark, Azure Data Factory, AWS Glue, Dataflow, Debezium, Fivetran, Airbyte, ClickHouse, PostgreSQL, SQL Server, Power BI, Tableau, Looker, LookML, QuickSight, Purview, LangChain, LangGraph, CrewAI, Qdrant, Milvus, Whisper, MLflow, FastAPI, Terraform, Kubernetes, Docker, GitHub Actions, Prometheus, Grafana, Python, SQL, Scala. 𝗔 𝗚𝗢𝗢𝗗 𝗙𝗜𝗧 𝗜𝗙 A data pipeline or warehouse that has outgrown how it was built. Disputed reporting needing one governed set of numbers. AI in your product needing a data foundation that survives it. Or an inherited estate nobody understands. 𝗣𝗥𝗢𝗕𝗔𝗕𝗟𝗬 𝗡𝗢𝗧 𝗔 𝗙𝗜𝗧 𝗜𝗙 A one-off dashboard with no underlying model, or a proof of concept nobody intends to run. Send me your setup and I will tell you straight whether it is worth the engineering.