LLM Orchestration Engineer

Datasirpi Pvt. Ltd.; Greater London; Hybrid; GBP 90,000 - 130,000; Full time

Job summary

DataSirpi is seeking an LLM Orchestration Engineer to lead the agentic infrastructure, designing systems where multiple AI agents collaborate to solve enterprise problems.

You'll move beyond prompt engineering into semantic memory, recursive reasoning loops, and multi-tool integration, ensuring our DataSirpi Agents are both powerful and predictable.

You are a software engineer first, and an AI enthusiast second, valuing testing, modularity and monitoring as core principles.

Qualifications

• Strong software engineering fundamentals including testing and modularity.
• Experience with AI agent orchestration and LLM systems.
• Familiarity with RAG pipelines and evaluation approaches.

Responsibilities

• Design and implement scalable agentic architectures using AutoGen and LangGraph.
• Optimize RAG pipelines for low-latency and high-relevance responses across massive datasets.
• Develop frameworks to measure LLM reliability, safety, and hallucination rates in production.
• Collaborate with Product and Data Engineering to bridge unstructured outputs with BI.

Skills

Software engineering

AI concepts

Systems design

Tools

Python

LangChain

AutoGen

Pinecone

Job description

At DataSirpi, we aren't just using AI; we're orchestrating it. As an LLM Orchestration Engineer, you will be the lead architect behind our agentic infrastructure, designing systems where multiple AI agents collaborate to solve high-stakes enterprise problems.

You'll move beyond simple prompt engineering into the realm of semantic memory, recursive reasoning loops, and multi-tool integration, ensuring our "DataSirpi Agents" are both powerful and predictable.

Technical Stack

Python

LangChain

AutoGen

Pinecone

Core Responsibilities

• Design and implement scalable agentic architectures using AutoGen and custom LangGraph implementations.
• Optimize RAG (Retrieval-Augmented Generation) pipelines for low-latency and high-relevance responses across massive datasets.
• Develop robust evaluation frameworks to measure LLM reliability, safety, and hallucination rates in production.
• Collaborate with Product and Data Engineering teams to bridge the gap between unstructured LLM outputs and structured business intelligence.

The Profile

You are a software engineer first, and an AI enthusiast second. You understand that "Agentic AI" is as much about traditional software engineering principles (testing, modularity, monitoring) as it is about neural networks.