Luca Holder | Data Science & AI

About Me

I'm an Associate Consultant for Cloud, Data & AI at b.telligent, currently delivering data platform and AI projects for manufacturing clients — after three years building a greenfield IoT data platform inside a Tier-1 automotive supplier. I work backwards from the customer's constraint, not forward from the tech stack: architecture workshops, hands-on Microsoft Fabric training, and production LLM systems built to hold up under real operational pressure. I'm finishing an M.Sc. in Data Science & AI alongside the day job.

Contact Me

Expertise

Data Engineering

Medallion & lakehouse architecture, dimensional modeling, and schema evolution — built with Apache Spark, Databricks, Airflow, Kafka, and Iceberg.

Machine Learning & Generative AI

Multi-tool AI agents, RAG pipelines, and prompt engineering — evaluated with golden datasets and LLM-as-judge. Shipped a production LLM agent stack for manufacturing anomaly detection.

Cloud Architecture

AWS (Lambda, API Gateway, DynamoDB, Cognito, S3, CloudFront) and Azure (Fabric, Synapse, Data Factory, IoT Hub, Purview) — defined as Infrastructure-as-Code with Terraform and shipped through CI/CD.

IT Consulting & Advisory

Architecture workshops, hands-on Microsoft Fabric training, and client enablement — so the teams I hand platforms over to can run them independently.

IoT & Manufacturing Data Platforms

Three years designing machine-data lake storage architecture and IoT platforms inside a Tier-1 automotive supplier, including data governance and BI tooling evaluation.

Databricks Certified Data Engineer Associate ITIL Foundation v4 Microsoft Security Operations Analyst Associate (SC-200) AWS Solutions Architect Associate (in progress) M.Sc. Data Science & AI — GoVersity (ongoing) B.Sc. Computer Science — DHBW Stuttgart

Projects

Cloud Architecture · AWS

Cloud-Native SaaS Platform on AWS

GitHub Profile ↗

Serverless application end-to-end: Amazon API Gateway (HTTP API) fronting an AWS Lambda FastAPI backend, DynamoDB as the data store, and Cognito for authentication.
S3 + CloudFront (origin access control) with ACM for a secured custom-domain frontend.
Entire stack defined as Infrastructure-as-Code with AWS SAM/CloudFormation, shipped through a GitHub Actions CI/CD pipeline that runs backend (pytest) and frontend (vitest) test suites on every merge, with a CloudWatch billing alarm guarding spend.

2026 — ongoing personal project

Generative AI · Production ML

LLM Agent Stack for Manufacturing Anomaly Detection

Put an LLM agent stack into production for manufacturing anomaly detection and natural-language querying.
Traced unreliable output to a grain mismatch and missing lineage in the source data, then rebuilt the data model so every alert carries a confidence threshold instead of firing on every deviation.
Delivered as part of four projects across three manufacturing clients at b.telligent.

Client work at b.telligent — implementation details limited by confidentiality

Data Engineering · IoT

IoT Data Platform & Governance

Authored the operating concept for taking a greenfield IoT data platform into productive use: end-to-end process chain, capacity planning, and cross-functional alignment.
Designed storage architecture for machine-data lakes from requirements through implementation, including a BI evaluation of SAP Analytics Cloud vs. Power BI.
Defined data-lake governance criteria and ran the Microsoft Purview evaluation covering cataloging, lineage, and access control.

ElringKlinger AG — Tier-1 automotive supplier, Oct 2022 – Sep 2025

Contact

Open to AI & data science collaborations, consulting discussions, research partnerships, and ambitious project ideas. Based in Gomadingen, Baden-Württemberg — open to relocation.

Send a Message