AI Agent Engineer

About the Role

We are a global digital agency comprised of strategists, creatives, media experts, data scientists, and engineers driven by one common purpose — accelerate business growth through marketing and digital transformation. Named a top 3% Google Premier Partner and recognized by Inc. 5000 and Adweek’s 75 Fastest Growing Companies, we’re constantly looking for “A” players to join our team.

The rapid growth is attributed to our strongest asset — our people. Our teams are highly collaborative and work closely with each client to set clear goals and objectives so that we can deliver exceptional results. Mindgruve is a place where every opinion is valued. Not only will you be empowered to contribute ideas, but you will also play a key role in the execution and driving success for brands across a variety of industries. Sound fun? Perfect — you’ll fit right in.

The AI Agent Engineer plays a foundational role in designing, building, and evolving our AI agent applications and supporting intelligence layer. This role sits at the intersection of LLM orchestration, backend engineering, cloud architecture, and analytics enablement. The primary responsibility is to architect and implement production-grade agent systems that can reason over business data, retrieve grounded context, maintain state, and stream useful outputs into internal and client-facing workflows.

This is a hybrid role—part application architect, part backend engineer, and part AI systems builder. You will work closely with Data Scientists, ML Operations Engineers, Data Engineers, and analytics stakeholders to create scalable, reliable, and well-governed agent experiences within an AWS-heavy environment. The work includes architecting streaming and asynchronous agent workflows, integrating with Bedrock and AgentCore services, connecting agents to knowledge bases and long-term memory, and ensuring that outputs are structured, testable, and enterprise-ready.

The role offers a unique opportunity to shape how AI agents are operationalized across a growing suite of analytics solutions. It is designed for someone who can move fluidly between system design and hands-on implementation, while helping establish best practices for RAG, evaluation, observability, and long-term maintainability.

What You'll Do Here

• Design, build, and maintain advanced AI agent applications using LangGraph, LangChain, and Haystack.
• Architect stateful, asynchronous, and streaming agent workflows, including event-loop-safe implementations using Python and real-time response streaming patterns.
• Design and maintain Retrieval-Augmented Generation (RAG) pipelines that connect agents to structured and unstructured business knowledge.
• Integrate agents with AWS Bedrock, Bedrock AgentCore, Bedrock Knowledge Bases, OpenSearch Serverless, S3 Vectors, Athena, and related AWS services.
• Define short-term and long-term memory strategies, including checkpointers, semantic retrieval, summarization flows, and user-context management.
• Implement strict output validation using Pydantic and structured schemas, including retry and self-correction workflows when LLM responses fail validation.
• Design prompt, tool, and routing patterns that allow agents to reason over analytics, marketing, media, and sales datasets in a grounded way.
• Collaborate with Data Scientists, MLOps Engineers, and analytics stakeholders to translate business needs into robust AI workflows and applications.
• Build maintainable APIs and backend services that expose agent functionality to internal tools and future productized interfaces.
• Establish testing patterns for agent applications, including mocked LLM calls, deterministic test coverage, and evaluation harnesses for regression prevention.

We Need a Person With

• Bachelor’s degree required in Computer Science, Data Science, Engineering, or a related field. Advanced degree is a plus.
• Strong proficiency in Python, including deep experience with backend application development.
• Hands-on experience building agentic applications with LangGraph required.
• Strong working knowledge of LangChain and Haystack required.
• Experience designing and implementing RAG systems, including retrieval strategies, chunking considerations, grounding patterns, and vector or hybrid search workflows.
• Experience working with AWS Bedrock and Bedrock AgentCore, including model orchestration and memory-aware agent design.
• Experience with structured LLM output design and validation using Pydantic or similar schema frameworks.
• Strong experience with API design, backend service development, and integration of AI systems into larger software ecosystems.
• Ability to work collaboratively across analytics, engineering, and product teams.
• Excellent verbal and written communication skills.
• Professional and personal integrity.
• A detail-oriented mindset focused on scalability, reliability, and production readiness.

What We Consider As a Plus

• Experience with Bedrock Knowledge Bases, OpenSearch Serverless, S3 Vectors, and memory strategies such as semantic, summarization, and user-preference memory patterns.
• Experience with pytest, unittest.mock, and test-driven development approaches for LLM-powered systems.
• Experience building agents that generate or evaluate SQL against Athena or related analytical query engines.
• Experience with Terraform or AWS CDK.
• Experience deploying AI applications within AWS-heavy environments and collaborating closely with MLOps and cloud infrastructure teams.
• Experience supporting analytics, marketing, or business intelligence use cases with AI systems.