Rohith Krishna Rangaraju

Lead Engineer · Almere, Netherlands · +31 647468395 · rangaraju963@gmail.com

rohithrangaraju.com · GitHub · LinkedIn · X

last updated 2026-04-28

Summary

Senior engineer with 8+ years across AI engineering, data engineering, and software engineering. Currently Lead Engineer at TomTom, shipping production multi-agent systems, custom-trained embeddings, MCP servers, and the data infrastructure that runs underneath. Strong technical-leader track record: mentoring engineers across all levels, translating ambiguous business requirements into shippable systems, and driving projects from concept to production deployment.

Core technical expertise

  • AI / LLM systems: production multi-agent systems, custom-trained embeddings, RAG pipelines, MCP servers, agentic UIs (Vercel AI SDK, Agno, AG-UI), prompt engineering, evals, fine-tuning small models for domain-specific tasks.
  • Backend & distributed systems: Python (expert), TypeScript, Node.js, REST APIs, FastAPI, Flask, gRPC-ready architectures, microservices, Server-Sent Events (SSE).
  • Data engineering: Spark Structured Streaming, PySpark, Apache Airflow, real-time ETL, medallion-architecture data lakes, Delta Lake, Azure EventHub.
  • Databases: PostgreSQL (incl. pgvector), Azure Data Explorer (Kusto), BigQuery, SQL Server, MySQL, semantic data modelling.
  • Cloud platforms: Azure (Databricks, AKS, OpenAI, Functions, ADLS, Key Vault), GCP (BigQuery, Cloud Composer, DataProc), AWS (S3, EC2).
  • Full-stack & DevOps: React, Next.js, TypeScript, Tailwind, GitHub Actions, Azure DevOps, MLFlow, Docker, Kubernetes.

Professional experience

Lead Engineer · TomTom · Amsterdam, Netherlands

January 2022 – Present

  • Lead and mentor a cross-functional team; team members promoted under my mentorship. Established AI/ML deployment best practices across three teams.
  • Architected the TomTom Traffic Agent, a production multi-agent AI system that answers plain-language traffic-analytics questions for strategic customers. Demoed by Product Management at CES 2026 Las Vegas.
  • Designed and built the multi-country geospatial similarity engine end-to-end: from a single anchor location it returns structurally-similar places across countries. Validated cross-country generalisation quantitatively (POI-density R² > 0.9 per country) before production.
  • Designed and built Query Intelligence, a natural-language interface for navigation that turns customer requests into grounded, deterministic execution plans. Iterated against real customer feedback through forward-deployed engineering.
  • Authored two open-source Model Context Protocol servers giving AI agents structured access to TomTom location and traffic-analytics services. Coordinated the Microsoft partnership for Azure Marketplace publication.
  • Architected the Data Quality Platform end-to-end (requirements + conceptual + technical design): a multi-tenant platform that lets every data team validate their pipelines without rolling their own framework. Implementation by junior engineers under my design oversight. Centralises contracts, not data.
  • Refactored the vehicle-telemetry silver layer, the production data layer that processes navigation telemetry from millions of in-car clients. Designed for query simplicity: PMs and ops engineers answer regional KPIs with a single SELECT instead of writing custom parsing code each time.
  • Built the developer-portal analytics APIs, the data-product layer that powers live customer-facing analytics on developer.tomtom.com. Every chart on the public dashboard renders from an endpoint I built.

Entrepreneurial ventures

Co-Founder & Technical Lead · AmplifyAI · amplifyai.us

2025 – Present (concurrent)

  • End-to-end GenAI platform automating innovation workflows for Consumer Packaged Goods companies, idea prototype to production-ready concept. React frontend, Python backend, multi-step LLM orchestration with deterministic post-processing where structure is required.

Co-Founder & Lead Developer · TheNewsBid · thenewsbid.com

2023 – 2024 (concurrent)

  • News-aggregation platform synthesising stories from many publishers into a balanced, citation-rich article. Removes single-source bias by design.
  • RAG-based study chatbot for India's public-service-exam students: production embeddings pipeline, vector search, conversational interface with retrieval-grounded citations.

Earlier experience

Senior Data Engineer · Oyo Vacation Homes · Eindhoven, Netherlands

May 2021 – December 2021

  • Architected an enterprise data lake on GCP (BigQuery, Cloud Storage, Apache Airflow / Cloud Composer) from a near-greenfield starting point.
  • Built a metadata-driven ETL framework in PySpark that cut new-pipeline development time by ~70%. Established schema-versioning and quality-validation patterns the team still uses.
  • Led and mentored the data-engineering team; ran knowledge-sharing sessions to elevate the team's technical capabilities.

Machine Learning Engineer · Tata Consultancy Services · Amsterdam, Netherlands

October 2019 – April 2021 · Client: ABN AMRO Bank

  • Led an ML team of 3 building a production NER + content-classification system for ABN AMRO, processing thousands of customer emails daily at 95%+ accuracy.
  • End-to-end MLOps with Azure Data Factory, Databricks, MLFlow: experiment tracking, model versioning, deliberate staging→production promotion. Implemented Delta Lake for data versioning and SCD handling.
  • DIME 2.0: lead engineer on a cloud-native distributed data-quality platform with a pluggable connector framework (SQL Server, BigQuery, Synapse, Postgres, Delta Lake, S3 / ADLS). Adopted across multiple enterprise data-lake deployments.
  • ADBConnectors: PySpark-based Python package abstracting I/O between Azure Databricks and common enterprise sinks (Synapse, SQL Server, Cosmos DB, blob, JDBC). Open-sourced internally within ABN AMRO so other teams could pick it up.

Data Engineer · Tata Consultancy Services · Hyderabad, India

October 2018 – October 2019 · Client: Citi Bank Singapore

  • Optimised long-running ETL for Citi's machine-learning use cases: cut Data Stage job execution from 4 hours to 1 hour through Python multiprocessing + PySpark parallelisation.
  • Designed real-time data handling with Apache Kafka and Python; introduced AVRO / Parquet formats and compression strategies to reduce HDFS storage cost.
  • Mentored 2-3 junior engineers on distributed-computing patterns and performance debugging.

Python Developer · Tata Consultancy Services · Bengaluru, India

June 2017 – October 2018 · Client: Sainsbury's (UK's largest retailer)

  • Built the user-management platform for Sainsbury's in Python Flask: REST APIs, OOP design patterns, Agile SDLC.
  • Worked directly with Sainsbury's Corporate Pingit Business Product Owners on change delivery and platform support.

Open source

Publications

  • MUTAWEB: Mutation Testing Tool for Servlet-based Web Applications, CVR Journal of Science and Technology, Vol. 13 (2017).

Education

Bachelor of Science in Computer Science, CVR College of Engineering (JNTU), India · 2013 – 2017.