New opportunity

Forward Deployed AI Development Expert

SAP · Singapore

About this role

SAP seeks a Forward Deployed AI Development Expert to provide technical leadership for strategic customer AI engagements. The role, described as a Principal Forward Deployed Architect, owns enterprise AI architecture from discovery through production, leads multidisciplinary delivery teams, and works with executive stakeholders on AI transformation programs. The position is based in Singapore, is marked hybrid, and involves 0–10% expected travel.

Responsibilities

Architect solutions spanning AI agents, RAG, knowledge graphs, MCP, multi-agent systems, enterprise integrations, and cloud-native platforms. Build scalable platforms using SAP AI Core, SAP BTP, Joule, Kubernetes, hyperscaler services, and SAP integration technologies.
Act as the primary technical advisor to customer executives, enterprise architects, and program leaders. Lead discovery workshops, architecture reviews, and executive solution presentations.
Establish reference architectures, reusable frameworks, engineering standards, governance models, and delivery patterns. Guide teams across customer engagements and mentor senior engineers and architects.
Address security, observability, performance, compliance, and operational excellence. Share implementation feedback with SAP product teams and develop new capabilities for Forward Deployed Engineering.

Requirements

Bachelor’s or master’s degree in Computer Science, Artificial Intelligence, Engineering, or a related discipline; at least 10 years of software engineering and enterprise architecture experience, including significant customer-facing delivery.
Expertise in Python and enterprise application architecture; deep experience with LLMs, agentic AI, RAG, GraphRAG, knowledge graphs, vector databases, MCP, tool calling, evaluation frameworks, and AI observability.
Extensive SAP BTP, SAP AI Core, Joule, SAP integration, and enterprise landscape experience. Strong background in Kubernetes, Docker, CI/CD, DevSecOps, AWS, Azure, GCP, and distributed systems; ability to integrate SAP and non-SAP systems at scale.
Experience leading cross-functional teams, mentoring senior technical staff, communicating with executives, and balancing business outcomes with technical and delivery needs.

Preferred

Experience defining enterprise AI strategy or AI Centers of Excellence; patents, publications, conference speaking, or open-source contributions; reusable platforms or engineering frameworks; and knowledge of regulated industries and enterprise governance.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

Enterprise AI architectureSolution architecturePythonAI agentsMulti-agent systemsLarge language modelsRAGGraphRAGKnowledge graphsVector databasesModel Context Protocol (MCP)Tool callingAI evaluationAI observabilitySAP AI CoreSAP BTPJouleSAP integration technologiesKubernetesDockerCI/CDDevSecOpsAWSAzureGoogle Cloud PlatformDistributed systemsEnterprise integrationSecurityTechnical leadershipExecutive stakeholder management

Checking your profile…

Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.

Understand the role

What matters most

Own enterprise AI architecture through production
The role owns architecture from discovery through production, spanning agents, RAG, knowledge graphs, MCP, multi-agent systems, enterprise integrations, and cloud-native platforms.
Lead strategic customer engagements
Serve as the primary technical advisor to customer executives, enterprise architects, and program leaders; lead discovery workshops, architecture reviews, and executive presentations.
Build scalable SAP-centered platforms
Architect platforms using SAP AI Core, SAP BTP, Joule, Kubernetes, hyperscalers, and integration services, including connections between SAP and non-SAP systems.
Set standards across engagements
Define reference architectures, reusable frameworks, golden paths, engineering standards, and governance models while guiding delivery teams and mentoring senior technical staff.
Make AI solutions production-ready
Drive security, observability, performance, compliance, and operational excellence; relay architectural feedback from customer implementations to SAP product teams.

Likely challenges to think through

Choosing architecture without needless complexity
Preparation hypothesis: The breadth of agents, RAG, graphs, and MCP could make it difficult to choose the simplest viable design. Prepare to explain when each approach earns its complexity and how evaluation evidence would inform the choice.
Balancing customer outcomes with production controls
Preparation hypothesis: Executive goals and delivery pace may compete with security, compliance, performance, and operational needs. Prepare a decision framework that makes those trade-offs visible from discovery through production.
Reusing patterns without flattening customer differences
Preparation hypothesis: Standards intended to accelerate multiple engagements may not fit every enterprise landscape. Prepare to distinguish what belongs in a reusable golden path from what should remain specific to a customer integration or governance context.

Practice your answers

Technical practice

How would you take an enterprise AI architecture from discovery through production?

What to coverThe role owns architecture across that lifecycle. Preparation suggestion: explain how you would turn business outcomes and enterprise constraints into design decisions, then validate those decisions before production.

When would you choose RAG, GraphRAG, or a knowledge graph?

What to coverThe posting names all three approaches, along with vector databases. Preparation suggestion: compare the questions each approach handles well, its data preparation and retrieval trade-offs, and how you would test answer quality.

How would you design a controlled agentic workflow using tool calling and MCP?

What to coverThe role covers AI agents, multi-agent systems, MCP, and tool calling. Preparation suggestion: discuss when multiple agents add value, how tools receive appropriate permissions, and how you would handle failed or unsafe actions.

How would you divide responsibilities across SAP AI Core, SAP BTP, and Joule in an AI solution?

What to coverThe posting calls for scalable platforms using these SAP technologies. Preparation suggestion: describe a plausible division of responsibilities, identify assumptions that depend on the customer's landscape, and explain how you would verify the fit.

How would you integrate an AI solution with SAP and non-SAP systems at scale?

What to coverLarge-scale enterprise integration is an explicit part of the role. Preparation suggestion: address data access, interface boundaries, authorization, failure handling, and the trade-off between direct integration and reusable services.

How would you deploy and evolve an AI platform across cloud-native environments?

What to coverThe posting names Kubernetes, Docker, CI/CD, DevSecOps, distributed systems, and AWS, Azure, and GCP. Preparation suggestion: explain deployment and rollback controls, environment portability, and where standardization may conflict with customer-specific needs.

How would you build security and compliance into an enterprise AI architecture?

What to coverThe role explicitly calls for security, compliance, and governance. Preparation suggestion: cover access boundaries, protection of enterprise data, auditability, and how controls would be reviewed as an AI solution moves into production.

How would you evaluate and observe an LLM-based enterprise application?

What to coverEvaluation frameworks and AI observability appear in the requirements. Preparation suggestion: distinguish pre-release evaluation from production monitoring, including answer quality, tool behavior, failures, and signals that warrant investigation.

How would you diagnose poor performance in a production RAG or agent workflow?

What to coverPerformance and operational excellence are stated responsibilities. Preparation suggestion: trace retrieval, model, tool, and integration steps separately; explain how you would weigh latency, reliability, quality, and cost when choosing a remedy.

What belongs in a reusable reference architecture or golden path for enterprise AI?

What to coverThe role defines reference architectures, frameworks, golden paths, and engineering standards across engagements. Preparation suggestion: distinguish reusable defaults from customer-specific decisions and explain how implementation feedback would improve the pattern.

Behavioural practice

Tell me about a time you advised executives on a consequential technical trade-off.

What to coverThe role serves as a technical advisor to customer executives. Preparation suggestion: use a real situation, your actions to explain options and business implications, and the resulting decision or outcome.

Describe a discovery workshop that changed your understanding of a customer's problem.

What to coverLeading technical discovery workshops is a stated responsibility. Preparation suggestion: give the situation, the questions and actions that uncovered a constraint, and how the resulting architecture or delivery plan changed.

Tell me about a time you led a cross-functional team while mentoring senior technical colleagues.

What to coverThe posting emphasizes multidisciplinary delivery and mentoring senior engineers and architects. Preparation suggestion: describe the situation, how you established direction without taking over every decision, and the observable result.

Describe a time you established an engineering standard across multiple teams or engagements.

What to coverThe role calls for reusable patterns and governance across engagements. Preparation suggestion: explain the situation, how you earned adoption while allowing justified exceptions, and what changed as a result.

Tell me about a time implementation feedback influenced a product or platform decision.

What to coverThe role includes sharing architectural feedback from customer implementations with SAP product teams. Preparation suggestion: use a real situation, show how you turned delivery evidence into actionable feedback, and state the result without claiming influence you cannot substantiate.

Sharpen your knowledge

Topics to prepare

Enterprise AI architecture patterns
Posting: The role architects LLM, agentic AI, RAG, GraphRAG, knowledge graph, vector database, MCP and tool-calling solutions. Revision exercise: compare when each pattern is appropriate in an enterprise solution, including the trade-offs of a multi-agent design.
SAP AI and enterprise landscape
Posting: Extensive experience with SAP BTP, SAP AI Core, Joule and SAP integration technologies is requested. Revision exercise: explain the architectural responsibilities of each and how an AI solution would connect SAP and non-SAP systems at scale.
Cloud-native and distributed architecture
Posting: The role calls for Kubernetes, Docker, CI/CD, DevSecOps, AWS, Azure, GCP and distributed-systems knowledge. Revision exercise: review the trade-offs in deploying and operating one scalable AI service across these architectural concerns; the posting does not prescribe a particular cloud.
Enterprise AI governance and operations
Posting: The architect drives security, observability, performance, compliance and operational excellence. Revision exercise: define the decisions and review criteria you would include in a governance model for an AI solution moving from discovery to production.
Reusable architecture and customer leadership
Posting: The role defines reference architectures, golden paths and engineering standards while advising executives and guiding delivery teams. Revision exercise: distinguish which decisions belong in a reusable standard and which should remain specific to a customer engagement.

Skills to revise

Python AI integration
Posting: Python expertise and experience with MCP and tool calling are requested. Suggested practice: sketch a Python-based tool-calling flow, including input validation, error handling and the boundary to an enterprise system.
SAP and non-SAP solution design
Posting: The role architects large-scale integrations and platforms using SAP AI Core, SAP BTP, Joule and enterprise integration services. Suggested practice: draw a discovery-to-production architecture showing component responsibilities, integration boundaries and key design decisions.
Evaluation and observability planning
Posting: Evaluation frameworks and AI observability are explicit experience areas. Suggested practice: draft an evaluation plan for a RAG or agentic solution, specifying what to test before release and what to monitor in production.
Cloud-native delivery review
Posting: Kubernetes, Docker, CI/CD and DevSecOps are requested. Suggested practice: walk through a deployment and review plan for an AI service, covering release controls, security, performance and operational handoff.
Executive architecture communication
Posting: The architect leads discovery workshops, architecture reviews and executive presentations. Suggested practice: prepare a short solution presentation that connects business outcomes to architectural trade-offs, then rehearse responses to delivery and governance questions.

Your preparation checklist

Map the stated requirements to evidence
Prepare a one-page evidence map for the posted degree, 10+ years of engineering and architecture experience, customer-facing delivery, Python, SAP platforms, cloud infrastructure, and technical leadership. Note gaps without claiming experience you cannot substantiate.
Build an end-to-end engagement walkthrough
Select a real customer-facing architecture project and outline discovery, decisions, delivery, production operations, business outcome, and your contribution. The role calls for ownership from discovery through production.
Practice a SAP-centered AI design exercise
Sketch a hypothetical architecture connecting SAP and non-SAP systems through SAP BTP and AI Core, with a clear role for Joule where appropriate. Explain when you would use RAG, a knowledge graph, MCP, tool calling, or multiple agents—and when you would not.
Prepare an operational review of that design
Create a short checklist covering security, compliance, evaluation, observability, performance, Kubernetes deployment, and CI/CD or DevSecOps. These are stated concerns; your proposed controls are preparation ideas, not a specified SAP design.
Show how you make architecture reusable
Bring an example, if available, of a reference architecture, framework, golden path, or engineering standard you helped establish. Explain adoption across teams and what you would change based on implementation feedback.
Rehearse discovery and executive communication
Draft workshop questions about business outcomes, enterprise landscape, constraints, and delivery ownership. Practice a brief executive explanation of architecture trade-offs, then prepare a leadership example involving cross-functional teams or mentoring.
Prepare targeted questions for SAP
Ask how this role coordinates architecture across customer engagements, turns implementation feedback into product input, and measures whether reusable patterns improve delivery. These are questions to investigate, not assumptions about SAP's processes.

Availability last checked: 4 Nov 2026. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in Singapore. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…