Hybrid AI Agents Engineer - Full-Stack on Azure

SoftServe; Abu Dhabi; Hybrid; AED 420,000 - 640,000; Full time

Job summary

SoftServe is seeking a senior full-stack engineer to design and build AI-powered applications on Microsoft Azure. You will own features end-to-end, from responsive front-end experiences to scalable back-end services and agent orchestration layers.

Role is hybrid, requiring office attendance at least twice a week. You will develop UIs for conversational experiences, real-time streaming responses, and monitoring dashboards, while building robust Python services and agent workflows on the Azure

Qualifications

• 8+ years of full stack engineering experience, with at least 3 years on the Azure stack.
• Strong backend skills in Python with distributed systems knowledge.
• Hands-on experience building applications with LLMs and agent frameworks (Semantic Kernel, AutoGen, LangGraph, or similar).
• Experience with Azure OpenAI, Azure AI Foundry, and vector databases (Azure AI Search, Cosmos DB vector search).
• Proven experience with RAG architectures, embeddings, chunking/indexing strategies, prompt orchestration, and tool/function calling.
• Hands-on experience with evaluation frameworks for LLM/agent outputs and defining quality metrics for AI systems.
• Experience designing end-to-end agentic solution architectures (multi-agent orchestration, tool/agent boundaries, error-handling and fallback strategies).
• Experience building real-time features (WebSockets, SignalR, SSE) for streaming AI/agent responses.
• Solid understanding of authentication/authorization (OAuth2, Entra ID) and DevOps practices (CI/CD, IaC, Docker/Kubernetes).

Responsibilities

• Design and build full stack, production-grade applications based on AI agents and multi-agent architectures.
• Develop modern, responsive web UIs supporting conversational experiences, real-time streaming agent responses, and dashboards for monitoring agents.
• Build robust back-end services and APIs (REST, gRPC, MCP) in Python (FastAPI), enabling agent-to-agent and agent-to-system communication.
• Develop agentic workflows using frameworks such as Semantic Kernel, AutoGen, LangChain/LangGraph, or Azure AI Agent Service.
• Integrate LLMs via Azure OpenAI Service, including prompt engineering, function/tool calling, and RAG pipelines.
• Design chunking and indexing strategies for RAG pipelines (semantic chunking, hybrid search, re-ranking) to optimize retrieval quality.
• Build and maintain evaluation pipelines for LLM/agent outputs (e.g., RAGAS, promptfoo, or custom eval frameworks) to measure accuracy, relevance, and reliability.
• Design cloud-native solutions using Azure Functions, AKS, Container Apps, Cosmos DB, and Azure AI Search, including data models across SQL and NoSQL stores.
• Implement secure, observable, cost-optimized systems using Azure Monitor, Key Vault, Entra ID, and CI/CD pipelines with infrastructure-as-code.

Skills

Full stack engineering

Python

Distributed systems

LLMs & agent frameworks

Azure expertise

Real-time features

Tools

Azure OpenAI

Cosmos DB vector search

Semantic Kernel

AutoGen

LangChain/LangGraph

Azure AI Search

Docker/Kubernetes

CI/CD

Infrastructure as Code

Job description

SoftServe is seeking a senior full-stack engineer to design and build AI-powered applications on Microsoft Azure. You will own features end-to-end, from responsive front-end experiences to scalable back-end services and agent orchestration layers.

Role is hybrid, requiring office attendance at least twice a week. You will develop UIs for conversational experiences, real-time streaming responses, and monitoring dashboards, while building robust Python services and agent workflows on the Azure