Agentic AI Principal (m/f/d)

## **Job responsibilities**

• *Shaping and winning**
• **Agentic AI advisory and strategy:** shape and solution complex German, Dutch and DACH pursuits with account, commercial and delivery leaders, translating client needs into differentiated propositions, credible architectures and executable delivery approaches that support regional pipeline conversion.
• Independently drive complex pursuits with technical and executive stakeholders, including shaping the commercial model (fixed-price, outcome-linked), trade-offs and risk.
• Act as a sparring partner to C-suite buyers (CEO, CIO, CDO, CISO, COO): frame business-outcome hypotheses, challenge assumptions and define a thin slice of agentic delivery that proves value quickly.
• Lead AI maturity assessments, identify and prioritise processes for agentisation, and connect architecture choices to measurable business value and ROI.
• Represent Thoughtworks' agentic AI expertise in strategic client conversations, industry forums, partnerships and thought leadership where appropriate.
• *Designing and building**
• **Design and build engineering:** set the architecture direction for complex, production-grade agentic AI systems, including multi-agent workflows, orchestration, tool use, memory, context strategy, retrieval, evaluation, observability and enterprise integration.
• Integrate agents with the German enterprise core: SAP (S/4HANA, BTP), LIMS, QMS and clinical systems, MES/PLM and industrial systems, and legacy back-office platforms, using open protocols such as MCP and A2A where they fit.
• Ensure agents run on trustworthy data foundations, working with data engineering and data product teams on data quality, semantics and access.
• Design solutions for regulated German and EU environments, with strong consideration of data sovereignty (EU and German sovereign cloud, open-weight or self-hosted models where required), privacy, security, responsible AI and applicable regulation, including the EU AI Act and GDPR; for healthcare and life sciences GxP, EU MDR/IVDR and the European Health Data Space; and for other sectors DORA, NIS2 and BaFin requirements where relevant.
• Guide teams in balancing rapid experimentation with robust production engineering, including evaluation-driven development, testing, CI/CD, reliability, cost and performance optimisation.
• Provide technical and delivery leadership for complex programmes, staying hands-on enough to validate architecture, unblock teams, manage technical risk and ensure successful execution.
• Partner with product, data, software engineering, enterprise architecture, governance and change specialists to design end-to-end solutions that deliver measurable business outcomes.
• *Governance and safeguards**
• Establish reusable patterns for observability, auditability, traceability and control of agentic systems, enabling teams to demonstrate safe and reliable behaviour in production.
• Translate EU AI Act obligations (risk classification, human oversight, logging, technical documentation, GPAI-provider responsibilities) into concrete engineering controls and evidence.
• Engage early with works councils (Betriebsrat, ondernemingsraad), data protection officers and information security, supporting DPIAs and co-determination processes so agentic solutions that touch employee workflows can go live.
• Anticipate technical, organisational and governance challenges in enterprise AI adoption and define practical approaches to address them.
• *Principal-level leadership and growing capability**
• Set reusable architecture patterns, engineering standards and governance practices across multiple teams and engagements, influencing the evolution of Thoughtworks' agentic AI capability beyond an individual programme.
• Set technical direction across multiple teams and initiatives, defining reusable patterns, frameworks and engineering standards for agentic AI delivery.
• Mentor senior engineers and technical leads, build the next generation of the German team, and raise AI fluency and agentic engineering standards across specialists and generalists.
• Evaluate emerging agentic AI technologies and frameworks and guide their responsible adoption based on client context, engineering quality and business value.
• Build our German and Dutch market presence through local-language points of view, ecosystem and partner initiatives (hyperscalers, SAP, European model providers) and community events.

## **Job qualifications**

### **Technical Skills**

• Deep expertise in modern software engineering and Python, with a strong track record of setting standards for testing, CI/CD, observability, maintainability and production reliability.
• Recognised experience architecting and delivering complex GenAI and agentic AI systems at enterprise scale, from early discovery through production deployment.
• Deep knowledge of agentic architectures and frameworks such as LangGraph, AutoGen, CrewAI, LlamaIndex or equivalent, including tool use, orchestration, memory, multi-agent coordination and interoperability protocols (MCP, A2A).
• Strong expertise in Retrieval-Augmented Generation, advanced information retrieval, vector databases and enterprise knowledge integration.
• Experience designing evaluation strategies for GenAI and agentic systems, including quality, safety, reliability, latency and cost measures.
• Strong understanding of AI governance and safeguards, including guardrails, red-teaming, observability, responsible AI and secure deployment practices (prompt injection, data leakage, tool misuse).
• Significant experience deploying AI systems on major cloud platforms using containers, orchestration and production-grade infrastructure patterns, including EU sovereign cloud, hybrid or on-premises setups and self-hosted open-weight models.
• Experience with model adaptation and optimisation techniques, including prompt/context engineering, fine-tuning where appropriate, and application-level performance optimisation.
• Ability to make sound architecture trade-offs across model capability, reliability, security, compliance, performance and cost.
• Working knowledge of the EU AI Act, GDPR and sector rules (for healthcare and life sciences: GxP and GAMP 5 computerised-system validation, EU MDR/IVDR, EHDS; for other sectors: DORA, NIS2, BaFin) and how they shape agentic system design.
• **Strongly preferred:** delivering AI in healthcare or life sciences (e.g. R&D, clinical, regulatory, manufacturing quality, commercial).
• Nice to have: integrating agents with SAP and industrial or manufacturing systems; data product and data platform foundations for AI.

### **Consulting and Commercial**

• Proven ability to lead C-suite and senior client engagements, rapidly establish credibility and advise on both business and technical choices.
• Strong consulting and commercial judgement: you can point to data and AI (ideally agentic or GenAI) pursuits you helped shape, sell and deliver in the last three to five years, from around €3M–5M upwards, including fixed-price or outcome-linked elements.
• Experience shaping, solutioning, selling and delivering complex deals of €5M+ in the last few years. Annual target of solutioning, selling and delivering €10M+ TCV in Data & AI deals with an AI-first delivery focus.

### **Professional Skills**

• Executive presence and communication skills, making complex agentic AI concepts, risks and trade-offs clear to technical and non-technical audiences.
• Proven leadership of complex delivery across multidisciplinary teams, moving fluidly between strategy, architecture, governance and hands-on technical leadership.
• Ability to influence across organisations without formal authority, align diverse stakeholders and navigate ambiguity in high-value or high-risk programmes.
• Proven ability to mentor senior engineers and technical leaders and raise standards across a wider engineering community.
• Self-directed, prioritisation-led working style, with creative problem-solving in fixed-scope, outcome-driven work.
• Experience navigating German stakeholder structures, including works councils, data protection officers and procurement.
• **German B2 minimum**, C1 preferred; fluent English

Wir zeigen einen Auszug und einen Link je Anzeige, und den vollen Text, wenn die Quelle es erlaubt. Genannte Marken und Logos gehören ihren Inhabern.

Quelle: Statistisches Bundesamt (Destatis), Tabelle; dl-de/by-2-0; eigene Berechnung

Veröffentlicht am 01. Nov. 2026