Carlos Hernandez Industrial AI and governance

Carlos Hernandez

AI Strategy Lead & Compliance Officer, Bosch Mobility Electronics

I help industrial organizations adopt AI they can deploy, govern, and trust.

I lead AI strategy and compliance for Bosch Mobility Electronics, moving manufacturing software, developer tooling, security, and governance from roadmap to deployment. GenAI Gurus, agent governance work, and the robotics notes below follow the same questions in public.

Professional work GitHub

Stuttgart region. Working across Europe.

Carlos Hernandez
40k employee organization context 20+ AI solutions in rollout 1,600 software developers in tooling scope ~3,000 GenAI Gurus members

How the work connects

One arc, three threads.

From enterprise AI strategy, through the developer ecosystem, to governable autonomy. The same questions run across my role at Bosch and the public projects below.

Professional work

AI strategy, governance, and developer enablement at industrial scale.

My current work is AI strategy and compliance for Bosch Mobility Electronics: portfolio direction, deployment paths, developer enablement, and governance for practical AI adoption.

Current role - Bosch Mobility Electronics

AI Strategy Lead & AI Compliance Officer

I own the AI strategy agenda: priorities, governance, portfolio direction, and quarterly management cadence for a 40,000-employee organization. I convert GenAI and industrial AI opportunities into adoption at that scale.

  • Orchestrate rollout of 20+ AI solutions, including agentic AI tooling for 1,600 software developers
  • Assess agentic AI use cases for security and compliance risk: tool access, human approval, telemetry, and evidence
  • Lead AI governance and compliance across a 2,000-application IT landscape, aligned to the EU AI Act
  • Secure funding and central financing for strategic AI use cases, and convert adoption assessments into executive actions
  • Lead an internal AI developer community for adoption, feedback loops, and best-practice sharing
LinkedIn profile
Role scope Strategy -> portfolio -> adoption -> governance
Organization 40k employees
Rollout 20+ AI solutions
Tooling 1,600 software developers
Governance 2,000 applications

This is internal operating-model and deployment work for Bosch. The writing and code below cover a related question instead: what happens when AI systems move from pilot into operation.

2021-2025

AI and data platform organization

Built and scaled a 30+ engineer AI and data platform organization across Europe and India, led a EUR 3M annual investment, and established GenAI evaluation frameworks for production-oriented use cases.

2018-2020

AIoT products and edge AI

Led industrial IoT products including Bosch Intelligent Vibration Analysis Sensor, Bosch's first product running an on-device CNN with TensorFlow Lite for Microcontrollers on ARM Cortex-M7.

2015-2018

Software product delivery

Led Bosch Transport Data Logger from concept to market release across embedded software, algorithms, mobile apps, cybersecurity, and distributed engineering.

Related public work

Public work on the same questions.

These public projects run alongside the Bosch work above: inspectable pieces on practical governance, developer enablement, agent controls, and physical AI.

Reference lab

Governed Embodied Agent Lab

A runnable reference environment for policy-gated robot actions: trusted context, fail-closed decisions, action-bound approval, restart continuity, one-time approval use, and a tamper-evident evidence chain.

Open repository Research map
Demo run for the Governed Embodied Agent Lab showing approval, restart, replay blocking, and evidence verification

Open source

Agent governance work

Merged work across the Microsoft Agent Governance Toolkit and multiple AgenTrust repositories: runtime protocols, approval flows, sandbox hardening, audit patterns, and embodied action evidence.

17 Microsoft AGT PRs 8 AgenTrust PRs
Microsoft PRs AgenTrust PRs
Merged Microsoft Agent Governance Toolkit pull request for an action-bound approval protocol

Community / writing

GenAI Gurus and field notes

GenAI Gurus is a practitioner community for applied generative AI, connecting practitioners, startups, researchers, and enterprise AI leaders across Europe. I publish notes on governance, agents, robotics, and physical AI at the two sites below.

~3,000 members 50+ mentoring sessions
Visit GenAI Gurus Read Field Notes
GenAI Gurus homepage showing the community positioning and public metrics

How I think about it

AI is moving from generating outputs to taking actions. Governance has to move from documents into runtime systems.

Copilots Guidelines
Agents Runtime controls
Embodied AI Verifiable evidence

Recent contributions

Recent public contributions.

  1. Action-bound approval protocol

    Proposed and advanced shared semantics for approvals bound to action digest, policy version, approver identity, durable pending state, and replay protection.

    View merged PR #2953

  2. Versioned approval transport

    Wired versioned webhook approval contracts into Agent Mesh governance flows.

    View merged PR #3100

  3. Sandbox denial logging

    Added denied-command logging through a hardened-image shim, so the audit trail captures blocked actions too.

    View merged PR #3019

  4. Industrial embodied-AI governance in AgenTrust

    Proposed an embodied action evidence profile and a relying-party verification API, plus industrial examples for preserving evidence across controller restarts and governing physical-agent risk.

    View AgenTrust PRs

Writing and research

Two publications, same questions.

Governance notes cover the control mechanics. Robotics notes cover the same controls where the stakes are physical.

Related places

The same work, in different formats.

GitHub Code, PRs, labs, curated research maps LinkedIn Professional history and current work GenAI Gurus Practitioner community and events Governance Technical field notes for AI builders Robotics Embodied-AI research notebook EU AI Act Curated implementation resource

Credentials

Certifications, education, and community roles.

Program, security, architecture, and cloud credentials behind the strategy, governance, and developer-ecosystem work above.

ProgramPMP, Project Management Professional
AI securityMicrosoft Azure Security Engineer Associate
SecurityCompTIA Security+
AI governanceBlueDot AI Governance & AI Alignment
ArchitectureTOGAF Enterprise Architecture Practitioner
CloudAzure AI Engineer, Azure & AWS Solutions Architect

Education

  • MSc, Wireless Communication SystemsUniversity of Aalborg, Denmark, 2013
  • Telecommunications Engineering (BSc + MSc equivalent)University of Oviedo, Spain, 2013
  • Granted patent, WO2017EP70617Industrial sensing

Community and recognition

  • Founder, GenAI Gurus~3,000-member AI practitioner community
  • Founder, TinyML GermanyEdge and embedded AI community, 2019 to 2024
  • OWASP GenAI Security ProjectLLM and agentic AI security
  • Lecturer in Generative AI, IEBSAWS Community Builder, DeepLearning.AI mentor

Languages: English (fluent), German (fluent), Spanish (native).

About and contact

About

My work spans industrial AI adoption, developer ecosystems, AI platform strategy, GenAI, edge AI, and practical governance: turning AI capability into systems you can deploy and operate.

The links below show the parts of that work you can inspect.

Connect on LinkedIn View GitHub profile