Martin Wimmer-Senior AI Solution Architect

Azure AI Foundry, GitHub Copilot (Agent Mode), Model Context Protocol (MCP), Kubernetes, CloudFoundry, Terraform, GitHub Actions, Langfuse, Python, Grafana

Objective: Accelerate enterprise-wide developer enablement and migration from GitLab/Jenkins to GitHub through secure CI/CD standards and automated, agentic developer support.

• Architected & deployed an enterprise-grade AI Support Agent integrated into GitHub Copilot via MCP, enabling developers to query legacy Confluence docs and Git repositories contextually.
• Designed & standardized secure, reusable GitHub Actions "Golden Path" templates, accelerating onboarding and ensuring compliance-by-design for delivery teams.
• Built and engineered robust data pipelines to establish a DevOps Maturity Model, monitoring platform adoption and migration KPIs via Grafana and Azure Monitor.
• Established LLM observability and evaluation frameworks utilizing Langfuse and Azure Monitor to optimize agent responses and control token costs.

Azure App Service, Azure AI Foundry, Neo4j, Langfuse, Terraform, GitHub Actions, Managed Identities

Objective: Design and implementation of a secure, highly available GenAI infrastructure with a strong focus on zero-secret authentication

• Designed a scalable web application architecture, including the migration from a local K3s environment to a fully managed Azure platform.
• Implemented a strict Identity and Access Management (IAM) model using Azure Managed Identities to enable passwordless authentication between App Service, Key Vault, Storage Accounts, and Azure AI Foundry.
• Automated the complete infrastructure lifecycle using Infrastructure as Code (IaC) with Terraform and GitHub Actions, including isolated state management.

Azure (Databricks, Data Factory, Functions, SQL), Terraform, Azure DevOps (GitOps), SAP Emarsys

Objective: Design and provisioning of a scalable Cloud and DevOps foundation to enable automated customer segmentation and integration into enterprise CRM systems.

• Architected and provisioned the foundational Azure Data platform via Terraform and GitOps, establishing a secure and standardized development environment for a 12-person engineering team.
• Designed an enterprise integration layer bridging Databricks analytics with SAP Emarsys via Azure Data Factory and event-driven Azure Functions for automated campaign synchronization.
• Established CI/CD pipelines and operational monitoring within Azure DevOps, ensuring resilient data delivery and fault-tolerant infrastructure deployments.

Enterprise AI Strategy, AI Governance, C-Level Advisory, Use-Case Prioritization

Objective: Translation and operationalization of the corporate group's AI strategy into a tailored, actionable roadmap for the reinsurance subsidiary.

• Authored the subsidiary's official AI Strategy, effectively adapting parent-group directives into specific, high-value AI initiatives for the reinsurance sector.
• Led stakeholder workshops and C-level alignment to identify, evaluate, and prioritize ROI-driven AI and GenAI use cases based on business impact.
• Established an AI Governance and Responsible AI framework, defining guidelines for secure, compliant, and risk-mitigated model adoption.

End-to-end architectural responsibility for the company's transition to AI-driven products, leading the technical roadmap from PoC to production-grade Kubernetes deployments.

Project 1: GenAI Chatbot & Search Optimization

Azure AI Foundry, LangGraph, Next.js, Hybrid Search

Objective: Modernize customer support via an AI-driven, scalable search and chat ecosystem.

• Architected an intelligent support assistant utilizing Azure AI Foundry and LangGraph for reliable, context-aware query resolution.
• Engineered a full-stack Next.js frontend with streaming responses, seamlessly integrated with a newly implemented relevance-based hybrid search (Azure AI Search).

Project 2: Legal AI Automation (Production System)

AKS, Argo Workflows, GitOps, Terraform, Multi-Agent LLMs

Objective: Automate high-cost legal advisory services via the company's first production AI product.

• Designed a multi-agent validation architecture (LangGraph) featuring a supervisor LLM to guarantee high accuracy and mitigate hallucinations in tax/payroll contexts.
• Built a fully automated, cloud-native document ingestion pipeline using Argo Workflows, deployed to Azure Kubernetes Service (AKS) via Terraform and GitOps.

Project 3: Cloud-Native ETL Architecture Evaluation

Azure Data Factory, Synapse, Medallion Architecture

Objective: Strategic evaluation of data ingestion frameworks for GenAI systems.

• Prototyped a Medallion-based data pipeline utilizing Azure Data Factory and Synapse Notebooks for unstructured document processing.
• Delivered a conclusive cost-benefit analysis that strategically pivoted the company away from ADF towards a more cost-efficient Kubernetes/Argo architecture for the final product.

Project 4: Predictive Maintenance PoC

Deep Learning, LSTMs, Grafana, Time-Series Analysis

Objective: Proactive prediction of server anomalies to enhance infrastructure reliability.

• Engineered LSTM/GRU-based deep learning models to analyze multivariate time-series telemetry in real-time, achieving 90% predictive accuracy.
• Provided the technical feasibility baseline used for executive decision-making regarding cloud migration vs. on-premise infrastructure.

Python, TensorFlow/Keras, XGBoost, LightGBM, 3D Calibration

Objective: Enhancement of 3D scanner color fidelity via machine learning regression models.

• Engineered and tuned gradient boosting ensembles (XGBoost, LightGBM) alongside neural networks to predict and correct color anomalies in 3D scan data.
• Executed physical hardware calibration to establish high-quality ground-truth datasets for model training.