Job VC
AI Developer
Technologies
Description
Project Overview
The project focuses on building a production-ready enterprise AI chatbot on SAP BTP Kyma, leveraging modern AI and backend technologies such as
SAP Joule, Node.js/TypeScript, NestJS, mastra AI, OpenSearch, and SAP HANA.
The platform emphasizes AI architecture, reliability, and scalability, enabling advanced RAG pipelines, agent-based capabilities, and secure enterprise integrations. Strong engineering practices, observability, and AI quality evaluation are key success factors.
Role Responsibilities
As an AI Developer, the consultant will:
Design and implement agentic AI capabilities, including tool/function calling, planning, memory/state handling, safety guardrails, and failure recovery
Build and optimize advanced RAG pipelines on OpenSearch (hybrid retrieval, embeddings, rerankers, query rewriting, retrieval evaluation)
Develop and maintain API and MCP integrations, including MCP servers/clients, capability manifests, and clean, versioned REST/WebSocket APIs
Create and iterate prompt strategies, including tool-aware prompting, structured outputs (JSON), and prompt governance
Establish AI quality evaluation and testing, using tools such as RAGAS and/or DeepEval, including regression testing and red-teaming
Implement secure and performant data access with SAP HANA using TypeORM
Ensure observability and performance monitoring using Dynatrace (latency, cost, quality, SLIs/SLOs)
Build and maintain CI/CD pipelines for Kyma/Kubernetes environments
Collaborate closely with product and architecture teams, document designs, and participate in technical reviews
Mandatory Skills and Experience — MUST HAVE !!!
Proven production AI adoption, including agentic or advanced RAG solutions with measurable impact
Strong experience with advanced RAG on OpenSearch (vector and hybrid search, indexing strategies, retrieval evaluation)
Experience with agent frameworks and patterns (multi-step planning, tool integration, memory/state, guardrails)
Hands-on experience with AI evaluation and testing (RAGAS, DeepEval, or similar)
Strong Node.js and TypeScript skills; solid API engineering (REST/WebSockets, OpenAPI, versioning)
NestJS experience in enterprise-grade services
Experience with MCP integrations in production or strong evidence of rapid ramp-up capability
Strong experience with SAP HANA and TypeORM
CI/CD experience for containerized applications on Kyma/Kubernetes
Dynatrace experience for observability and performance monitoring
Security-first mindset (OAuth2/OIDC, JWT, RBAC, secure secrets handling)
Ability to learn quickly and work within complex enterprise architectures
Fluent English
Nice to Have
SAP ecosystem knowledge; SAP Joule experience is a plus
Experience with mastra AI or similar AI orchestration frameworks
Advanced prompt engineering skills and prompt governance workflows
Experience integrating AI into web frontends (streaming responses, SSE, rate/cost controls)
Experience with multiple LLM providers and routing strategies
Knowledge of data privacy and compliance (e.g. GDPR-aware AI design)
Project Details
Start: ASAP
Duration: Until 31.10.2026 (extensions possible)
Estimated Effort: 960 hours
Location: Fully remote
Language: English
Candidate Location: EU
The project focuses on building a production-ready enterprise AI chatbot on SAP BTP Kyma, leveraging modern AI and backend technologies such as
SAP Joule, Node.js/TypeScript, NestJS, mastra AI, OpenSearch, and SAP HANA.
The platform emphasizes AI architecture, reliability, and scalability, enabling advanced RAG pipelines, agent-based capabilities, and secure enterprise integrations. Strong engineering practices, observability, and AI quality evaluation are key success factors.
Role Responsibilities
As an AI Developer, the consultant will:
Design and implement agentic AI capabilities, including tool/function calling, planning, memory/state handling, safety guardrails, and failure recovery
Build and optimize advanced RAG pipelines on OpenSearch (hybrid retrieval, embeddings, rerankers, query rewriting, retrieval evaluation)
Develop and maintain API and MCP integrations, including MCP servers/clients, capability manifests, and clean, versioned REST/WebSocket APIs
Create and iterate prompt strategies, including tool-aware prompting, structured outputs (JSON), and prompt governance
Establish AI quality evaluation and testing, using tools such as RAGAS and/or DeepEval, including regression testing and red-teaming
Implement secure and performant data access with SAP HANA using TypeORM
Ensure observability and performance monitoring using Dynatrace (latency, cost, quality, SLIs/SLOs)
Build and maintain CI/CD pipelines for Kyma/Kubernetes environments
Collaborate closely with product and architecture teams, document designs, and participate in technical reviews
Mandatory Skills and Experience — MUST HAVE !!!
Proven production AI adoption, including agentic or advanced RAG solutions with measurable impact
Strong experience with advanced RAG on OpenSearch (vector and hybrid search, indexing strategies, retrieval evaluation)
Experience with agent frameworks and patterns (multi-step planning, tool integration, memory/state, guardrails)
Hands-on experience with AI evaluation and testing (RAGAS, DeepEval, or similar)
Strong Node.js and TypeScript skills; solid API engineering (REST/WebSockets, OpenAPI, versioning)
NestJS experience in enterprise-grade services
Experience with MCP integrations in production or strong evidence of rapid ramp-up capability
Strong experience with SAP HANA and TypeORM
CI/CD experience for containerized applications on Kyma/Kubernetes
Dynatrace experience for observability and performance monitoring
Security-first mindset (OAuth2/OIDC, JWT, RBAC, secure secrets handling)
Ability to learn quickly and work within complex enterprise architectures
Fluent English
Nice to Have
SAP ecosystem knowledge; SAP Joule experience is a plus
Experience with mastra AI or similar AI orchestration frameworks
Advanced prompt engineering skills and prompt governance workflows
Experience integrating AI into web frontends (streaming responses, SSE, rate/cost controls)
Experience with multiple LLM providers and routing strategies
Knowledge of data privacy and compliance (e.g. GDPR-aware AI design)
Project Details
Start: ASAP
Duration: Until 31.10.2026 (extensions possible)
Estimated Effort: 960 hours
Location: Fully remote
Language: English
Candidate Location: EU