FSA261005P2A1: Senior AI Agent Developer (CAP, Python, LangGraph, SAP AI Core) - Remote
joyIT Recruiting GmbH
Ansprechpartner: joyIT Recruiting GmbH
Berlin, Deutschland
ITAI AgentsRAGMLflowOpenTelemetryLLMAPIsCcdsPythonNode.jsSAP HANASemantic HTMLSolution ArchitectureVector Markup LanguageLarge Language ModelsGenerative AISAP FioriSAP UI5Artificial Intelligence Markup Language (AIML)Api-Management
Beschreibung
Independent development and enhancement of AI-driven business solutions on a business technology platform through agent-based workflows, backend services, and integrations.
The project extends enterprise AI capabilities through scalable, production-ready solutions interacting with enterprise systems and business data.
Responsibilities
The external contractor will independently:
- Design and implement AI agent workflows using LangGraph and Python.
- Develop and enhance CAP-based backend services, including CDS models and OData APIs.
- Configure and integrate AI Core and Generative AI Hub capabilities within solution architectures.
- Build and maintain MCP-compatible tool interfaces and agent integrations.
- Develop data persistence and retrieval mechanisms using HANA Cloud.
- Implement AI workflow orchestration, intent-routing logic, and human-confirmation mechanisms.
- Create technical documentation including architecture decisions, solution designs, and interface specifications.
- Develop automated tests for agent workflows, backend services, and integrations.
- Configure observability, monitoring, experiment tracking, and performance measurement where required.
- Assist in integration of AI capabilities with UI5/Fiori-based applications.
- Deliver technical recommendations regarding AI workflow implementation, optimization, and scalability.
Deliverables:
- Implemented LangGraph-based agent workflows according to approved solution specifications.
- CAP service enhancements, including CDS artifacts and OData service definitions.
- AI Core integration components and configuration artifacts.
- MCP-compatible tool interfaces and integration endpoints.
- HANA Cloud data models and persistence components supporting delivered functionality.
- Technical design documentation and architecture decision records.
- Automated test suites and execution results.
- Deployment-ready source code committed to the client's approved repository.
- Monitoring and observability documentation where applicable.
- Knowledge transfer documentation covering delivered solution components.
- Acceptance criteria:
- The client verifies completion by reviewing delivered artifacts against approved requirements and backlog items.
Acceptance includes:
- Delivered functionality successfully performs the agreed business process or use case.
- Components comply with defined architecture and integration requirements.
- Code successfully passes agreed quality checks and automated tests.
- APIs and integrations function according to documented specifications.
- Technical documentation is complete and reflects delivered functionality.
- Deliverables are demonstrated and accepted in sprint reviews, solution walkthroughs, or designated acceptance sessions.
- Critical or high-severity defects are resolved before acceptance.
SLAs / KPIs (completion target / timing):
- Deliverables completed within agreed sprint or project milestones.
- Documentation and code artifacts delivered with functional components.
- Defects identified during acceptance testing addressed within agreed delivery timelines.
- Development completed according to agreed backlog priorities and release schedules.
- Solution deliverables meet agreed quality standards and testing requirements prior to handover.
Qualifications
Required experience:
- Strong experience developing AI-based applications using Python.
- Experience with LangGraph or comparable agent orchestration frameworks.
- Proven experience with a business technology platform.
- Experience developing CAP applications using Node.js.
- Strong knowledge of HANA Cloud data modelling and persistence concepts.
- Experience designing REST/OData-based service integrations.
- Experience integrating Large Language Models (LLMs) into enterprise applications.
- Experience implementing automated testing and code quality controls.
- Ability to create technical architecture and solution documentation.
- Proficiency in English.
Preferred experience:
- AI Core and Generative AI Hub.
- MCP (Model Context Protocol) implementations.
- AI observability and experiment-tracking solutions (e.g., MLflow).
- OpenTelemetry-based monitoring and tracing.
- UI5 and Fiori application development.
- Vector databases, RAG architectures, and semantic search solutions.
- API Management.
- Enterprise AI governance and Responsible AI practices.
Additional informations
Work duration: 12 October 2026 to 31 December 2026.
Estimated workload: fulltime, 320 hours estimated spend.
100% Remote.
Maximum pay rate: 75€/h.
Language: English.
Next Steps
If the application process provides an option to send an email or a comparable notification, please activate it so that we can remain in contact with you.
After submission, you will receive an application link for the next steps.
Please apply only if qualified and your rate expectations suit the project.
We look forward to receiving your application.
Your team at joyIT.