Backend Engineer – LLM Integration

Luxoft Germany; United States; Remote; USD 120,000 - 160,000; Full time

Job summary

DXC Luxoft seeks a Backend Engineer to join a joint product team delivering an AI-powered in-vehicle voice assistant. The role emphasizes Python development, FastAPI microservices, and Kafka-based event streams within an AKS/Azure environment.

You will implement LLM orchestration, manage Azure OpenAI integration, and ensure robust testing and deployment practices. English proficiency (B2+) is required, and automotive domain experience is a plus.

Qualifications

• 3+ years professional Python development including production web services.
• Experience with FastAPI (or Flask/Django) for REST API design.
• Familiarity with at least one LLM orchestration framework: LangChain, LangGraph, or LiteLLM.
• Apache Kafka — producer/consumer implementation; at-least-once vs exactly-once semantics.
• Docker and Kubernetes basics (deployment manifests, service configuration).
• English B2 or above.

Responsibilities

• Build and maintain FastAPI microservices that orchestrate LLM calls and agentic workflows.
• Implement Kafka producers/consumers for real-time data and AI triggers.
• Integrate with Azure OpenAI endpoints and manage prompt routing.
• Write unit, integration and contract tests; maintain code quality gates in Azure DevOps.
• Containerise services with Docker; deploy on AKS and manage Helm/Terraform configs.
• Participate in code reviews and contribute to architectural decisions.
• Monitor service health with Azure Monitor; triage production incidents.
• Document APIs (OpenAPI) and maintain ADRs for key decisions.

Skills

Python

REST APIs

LLM orchestration

English (B2)

Tools

FastAPI

Kafka

Docker

Kubernetes

Job description

Project description

Our client is advancing its in-vehicle voice assistant into an intelligent, AI-powered companion. Since 2024, they have been incorporating large-language-model capabilities (Azure OpenAI / ChatGPT) into vehicles equipped with the MIB3 infotainment system, with new E³-architecture models featuring enhanced voice functions from the factory. The AIME (AI Model Engine) backend program supports this development over a multi-year timeline, addressing natural-language dialogue, empathic communication, and the complete cloud-edge data pipeline. DXC Luxoft serves as the end-to-end delivery partner, collaborating closely with the client’s engineers within a joint product team on the Azure platform (AKS, Azure OpenAI, Managed Identity, Azure Monitor, Azure DevOps).

Responsibilities

• Backend Engineers in this stream own the integration layer between the AI/ML components and the vehicle/cloud communication backbone. The stack is Python-first, event-driven via Kafka, and exposed through FastAPI services running on AKS. You will work inside the joint product team, picking up S/M/L backlog tickets and shipping production-ready code.
• Build and maintain FastAPI microservices that orchestrate LLM calls, tool use, and agentic workflows using LangChain / LangGraph.
• Implement Kafka producers and consumers for real-time voice session data, telemetry, and event-driven AI triggers.
• Integrate with Azure OpenAI endpoints via LiteLLM abstraction; manage prompt versioning and A/B routing.
• Write unit, integration, and contract tests (pytest, Pact); maintain coverage and quality gates in Azure DevOps.
• Containerise services with Docker; deploy and operate on AKS; contribute to Helm chart and Terraform configuration.
• Participate in code reviews with Audi engineers and contribute to joint architectural decisions.
• Monitor service health using Azure Monitor / Application Insights; triage and resolve production incidents.
• Document APIs (OpenAPI) and maintain ADR entries for significant implementation decisions.

Must have

• 3+ years professional Python development, including production web services.
• Experience with FastAPI (or Flask/Django) for REST API design.
• Familiarity with at least one LLM orchestration framework: LangChain, LangGraph, or LiteLLM.
• Apache Kafka — producer/consumer implementation; at-least-once vs exactly-once semantics.
• Docker and Kubernetes basics (deployment manifests, service configuration).
• English B2 or above.

Nice to have

• Azure services: AKS, Azure OpenAI, Managed Identity, CosmosDB.
• Terraform or Bicep for infrastructure-as-code.
• Automotive domain or OEM project experience.
• Async Python patterns (asyncio, aiohttp).
• German language skills.