About Our Client
Our client is an innovative AI-powered brand analytics platform that helps businesses understand their brand perception, competitive landscape, and market positioning through advanced AI-driven analysis. They leverage multiple Large Language Models to deliver actionable insights that enable brands to optimize their visibility and competitive positioning in the market.
The platform runs a production multi-agent system on AWS, orchestrating multiple AI providers across signal generation, content, and research workflows, and is exposed to clients through their own MCP server.
What We're Looking For
We are seeking an AI Engineer to help shape the technical direction of the AI platform. You will design and evolve the agentic systems that power this solution - multi-agent orchestration, LLM pipelines, evals, and the cloud infrastructure that runs them.
You should be hands-on, opinionated about how AI systems should be built, and excited to set the bar for an engineering team that ships AI features fast and safely. We value engineers who actively use AI agents and automation tools in their own workflows and stay current with the field as it moves.
Key Responsibilities
1. Agentic AI Systems
Design, build, and evolve multi-agent systems and LLM-powered pipelines. You will:
*
Design agent topologies - planner/executor, supervisor/worker, reflection loops, human-in-the-loop
*
Build and extend agents using frameworks like LangGraph, Strands, and Agent SDKs
*
Evolve the agent harness - execution loop, tool dispatch, context management, sub-agent spawning, and sandboxing
*
Design agent memory, context management, and tool-calling patterns
*
Extend the MCP server with new tools and capabilities
*
Enforce structured outputs and validation across LLM boundaries
2. LLM Quality, Evals & Observability
Build the layer that lets the team ship LLM features with confidence. You will:
*
Design and grow the eval platform - golden datasets, regression suites, LLM-as-judge
*
Integrate observability and tracing across providers and prompt versions
*
Track cost, latency, and quality per prompt, model, and client
*
Build guardrails for prompt injection, PII, and output safety
*
Drive prompt engineering practice - versioning, A/B testing, platform overlays
3. Cloud & Data Infrastructure
Own the cloud substrate that runs the AI workloads. You will:
*
Architect and maintain AWS infrastructure (ECS Fargate, Lambda, Step Functions, EventBridge, S3, Athena, DynamoDB, Bedrock)
*
Build and operate data pipelines that move LLM outputs from generation to analytics
*
Manage containerized deployments, CI/CD, and infrastructure as code
*
Ensure reliability, observability, and cost efficiency across the platform
4. Security, Compliance & AI Safety
*
Implement IAM, encryption, and network security best practices
*
Manage secrets and audit logging
*
Enforce multi-tenant isolation across agents, prompts, and data
*
Defend against prompt injection, jailbreaks, and PII leakage
5. Technical Leadership
*
Set technical direction for the AI platform and drive architecture decisions
*
Make build-vs-buy calls on frameworks, providers, and tooling
*
Mentor engineers and raise the bar on prompts, evals, and agent design
*
Collaborate with product to translate business goals into AI system architecture
What Makes This Role Unique
You will work on a production multi-agent platform with real scale and real users - multi-specialist agent systems, LLM pipelines, and a multi-provider stack running on AWS. The interesting problems are not bootstrapping; they are deciding what good looks like and building the systems that get you there