New opportunity

Forward Deployed AI Developer

SAP · Singapore

About this role

As a Forward Deployed AI Engineer (T2) at SAP, you will independently build AI-native enterprise applications and work directly with customers to solve complex business problems. You will own technical workstreams, design production-grade AI solutions, and collaborate with architects to deliver scalable, secure, enterprise-ready AI applications. Responsibilities: Design and implement production-grade AI applications using modern software engineering practices. Develop AI agents using LLMs, RAG, MCP, tool calling, structured outputs, and multi-agent orchestration. Build integrations with SAP products, enterprise applications, APIs, databases, and external services. Deploy and operate AI applications on SAP AI Core, SAP BTP, Kubernetes, and public cloud platforms. Own customer-facing technical workstreams and translate business requirements into working AI solutions. Develop reusable frameworks, accelerators, and reference implementations. Perform code reviews and mentor Associate AI Developers. Optimize AI systems for performance, reliability, observability, and cost. Drive innovation through prototypes and emerging AI technologies. Qualifications: Bachelor's or Master's degree in Computer Science, Artificial Intelligence, Information Technology, or a related discipline. 4–8 years of professional software engineering experience. Strong Python programming skills with experience in Java, TypeScript, or JavaScript. Hands-on experience with LangGraph, LangChain, CrewAI, Semantic Kernel, AutoGen, or similar frameworks. Strong understanding of LLMs, prompt engineering, embeddings, vector databases, RAG, MCP, and agentic AI. Experience building REST APIs, microservices, and cloud-native applications. Hands-on experience with Docker, Kubernetes, Git, CI/CD, and AWS, Azure, or GCP. Experience integrating enterprise applications and databases. Excellent communication, analytical thinking, ownership, and customer engagement skills. Preferred: Experience with Knowledge Graphs, GraphRAG, AI evaluation, and AI observability. Experience delivering enterprise AI projects in customer-facing roles. Exposure to SAP BTP, SAP AI Core, Joule, or SAP Business AI services is highly desirable.

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

Skills for this role

PythonJavaTypeScriptJavaScriptLangGraphLangChainCrewAISemantic KernelAutoGenLLMsPrompt EngineeringEmbeddingsVector DatabasesRAGMCPAgentic AIREST APIsMicroservicesCloud-native applicationsDockerKubernetesGitCI/CDAWSAzureGCPSAP BTPSAP AI Core

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

Production-Grade AI Development
The role requires moving beyond prototypes to build scalable, secure, and reliable AI-native applications. You must demonstrate proficiency in modern software engineering practices, including CI/CD, observability, and performance optimization, specifically within the SAP AI Core and BTP ecosystems.
Customer-Facing Technical Ownership
As a forward-deployed engineer, you will act as the primary technical lead for customer workstreams. You must be capable of translating ambiguous business requirements into concrete AI solutions while managing stakeholder expectations and collaborating directly with architects.
Agentic AI Orchestration
You are expected to design complex AI agents using frameworks like LangGraph, CrewAI, or AutoGen. This involves mastering tool calling, RAG, multi-agent orchestration, and structured outputs to solve real-world enterprise business problems.

Likely challenges to think through

Balancing Innovation with Enterprise Constraints
You will face the challenge of integrating cutting-edge AI technologies into rigid, secure enterprise environments. Prepare by articulating how you maintain system reliability and security while implementing experimental features like LLM-based agents.
Translating Business Needs to Technical Architecture
Customers often have complex, non-technical business problems. You must be prepared to demonstrate your process for requirements gathering and how you map those needs to specific AI patterns, such as RAG or Knowledge Graphs, without over-engineering.

Practice your answers

Technical practice

Designing Multi-Agent Orchestration

What to coverExplain how you would design a multi-agent system using frameworks like LangGraph or CrewAI to solve a complex business workflow. Discuss how you handle task delegation, state management between agents, and ensure consistent structured outputs.

RAG Implementation and Optimization

What to coverDescribe your approach to building a production-grade RAG pipeline. Focus on how you handle document chunking, embedding selection, and retrieval strategies to minimize hallucinations and improve accuracy in an enterprise context.

Integrating AI with Enterprise Systems

What to coverHow do you securely integrate LLM-based applications with existing SAP or legacy enterprise databases and APIs? Address authentication, data privacy, and the use of tools or MCP to bridge the gap between AI agents and structured business data.

AI Observability and Evaluation

What to coverWhat metrics and tools do you use to monitor the performance and reliability of AI applications in production? Explain how you implement evaluation frameworks to track drift, latency, and cost-efficiency over time.

Cloud-Native Deployment Strategies

What to coverDiscuss your experience deploying AI models and services on Kubernetes and cloud platforms like AWS, Azure, or GCP. Explain how you manage CI/CD pipelines for AI-native applications to ensure seamless updates and scalability.

Handling Tool Calling and Function Execution

What to coverWhen building agents that interact with external services, how do you ensure the model correctly selects and executes the right tools? Discuss your strategy for validating tool inputs and handling errors during the execution phase.

GraphRAG vs Standard RAG

What to coverCompare the use of Knowledge Graphs in RAG (GraphRAG) versus traditional vector-based RAG. When would you recommend one over the other for an enterprise application requiring high factual accuracy?

Scalability and Performance Tuning

What to coverHow do you optimize an AI application for high-throughput enterprise use cases? Discuss trade-offs between model size, inference latency, and cost when selecting LLMs for specific business tasks.

Behavioural practice

Translating Ambiguous Business Requirements

What to coverDescribe a time you had to translate vague or complex business requirements into a concrete technical AI solution. Focus on how you engaged with stakeholders to define the scope and ensure the final delivery met their needs.

Managing Technical Ownership

What to coverGive an example of a project where you owned a technical workstream from design to deployment. How did you handle obstacles, collaborate with architects, and ensure the solution was enterprise-ready?

Mentoring and Code Quality

What to coverDescribe your approach to performing code reviews and mentoring junior developers. How do you balance maintaining high engineering standards with the need to drive innovation and speed in AI development?

Handling Project Setbacks

What to coverTell me about a time an AI solution did not perform as expected in a customer-facing environment. What steps did you take to troubleshoot the issue, communicate with the customer, and pivot to a successful outcome?

Sharpen your knowledge

Topics to prepare

Multi-Agent Orchestration
The role requires building AI agents using frameworks like LangGraph or CrewAI. Exercise: Design a workflow where two specialized agents (e.g., a researcher and a summarizer) interact to solve a multi-step business problem, ensuring structured output and error handling. This is critical for the requirement to build complex, production-grade AI-native applications.
Enterprise RAG and Vector Databases
The position emphasizes RAG and vector databases for enterprise applications. Exercise: Implement a RAG pipeline that retrieves context from a mock enterprise database and uses an LLM to generate grounded answers. Focus on optimizing retrieval accuracy and handling document chunks, which is essential for the 'enterprise-ready' requirement.
Cloud-Native AI Deployment
The role involves deploying on SAP AI Core, BTP, and Kubernetes. Exercise: Create a containerized AI service using Docker and define a Kubernetes deployment manifest that includes resource limits and health checks. This directly supports the requirement to operate scalable, production-grade AI systems.

Skills to revise

Python for AI Engineering
Strong Python skills are mandatory. Exercise: Refactor a script that interacts with an LLM API to use asynchronous programming and robust error handling for API timeouts. This is necessary for building reliable, production-grade applications as requested.
Model Context Protocol (MCP) Implementation
The job explicitly lists MCP as a core requirement. Exercise: Build a simple MCP server that exposes a local database or file system as a tool for an AI agent to consume. This demonstrates your ability to integrate external services into AI workflows.
AI Observability and Evaluation
The role prefers experience in AI observability. Exercise: Integrate a logging or tracing tool (such as LangSmith or similar) into an existing agentic workflow to track latency, token usage, and output quality. This is relevant for the requirement to optimize AI systems for performance and reliability.

Your preparation checklist

Prepare a project walkthrough of an agentic AI solution
Develop a concise presentation or document detailing an AI application you built using frameworks like LangChain or CrewAI. Focus on how you implemented RAG, tool calling, or multi-agent orchestration to solve a specific business problem, highlighting your technical decision-making process.
Demonstrate cloud-native deployment expertise
Prepare to discuss your experience deploying containerized applications on Kubernetes or public cloud platforms (AWS, Azure, or GCP). Be ready to explain how you managed CI/CD pipelines, ensured system observability, and optimized for performance and cost in a production environment.
Articulate your approach to customer-facing technical work
Reflect on past experiences where you translated complex business requirements into technical AI solutions. Prepare specific examples of how you communicated with non-technical stakeholders, managed project scope, and balanced customer needs with enterprise-grade security and scalability requirements.
Formulate questions about SAP AI ecosystem integration
Prepare thoughtful questions regarding how the team leverages SAP BTP and SAP AI Core for enterprise deployments. Ask about the current challenges in integrating custom AI agents with existing SAP ERP data and how the team approaches AI evaluation and observability in a production setting.

Availability last checked: 18 Oct 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…