AI Engineer

Kobie Marketing; Richmond (VA); On-site; USD 100,000 - 130,000; Full time

Job summary

Medium is seeking an AI Engineer in Richmond, Virginia to build and deploy innovative agent solutions leveraging technologies like LangChain and Amazon AgentCore. This hands-on role entails collaborating closely with teams while ensuring reliability and quality of production code.

The successful candidate will have extensive experience in Python and a strong track record with LLMs, overseeing everything from development to operational success.

Qualifications

• 3+ years of professional Python experience.
• 1+ years of hands-on experience with LLMs in production.
• Working knowledge of LangChain/LangGraph.

Responsibilities

• Build agent harnesses in Python with tool-calling.
• Own and deploy prototypes through Amazon AgentCore.
• Partner with data engineers on retrieval patterns.

Skills

Python

LLMs

LangChain

Docker

SQL

Tools

Amazon CloudWatch

MLflow

Job description

Role Overview

We’re looking for a hands‑on AI Engineer to ship on our platform: building agent harnesses, writing the tools those agents call, and owning the reliability and evaluation of what goes to production. This is not a research role. You’ll prototype, ship, monitor, and iterate on features used by real teams.

Our team tends to be people who reason carefully, ship working code, and pick up new tools without a lot of handholding. There’s no single path into this role. We value the impact of what you’ve built and your track record of building things that hold up.

About the team and what we’ll build together

Kobie runs some of the largest loyalty programs in the world. We’re building an internal agent platform on Amazon AgentCore that automates analyst workflows, surfaces insights from program data in Snowflake, and gives our teams and clients an LLM‑native way to work with complex loyalty logic.

Role & Responsibilities – How you will make an impact

Agent Development

• Build agent harnesses in Python using LangChain and LangGraph, including tool‑calling, structured outputs (Pydantic/JSON schema), retries, streaming, and memory
• Package agent harnesses for the AgentCore Runtime with appropriate context, tools, skills, and subagents that fit cleanly into production flows and scenarios
• Write the tools and skills agents use: API integrations, SQL queries against Snowflake, and Snowflake‑backed knowledge retrieval with clear contracts and Pydantic validation

Evaluation and Reliability

• Build evaluation harnesses (golden datasets, LLM‑as‑judge, regression suites) using AgentCore Evaluations, and wire them into CI
• Implement guardrails around tool execution: auth scoping, input/output validation, PII and prompt‑injection protections, and hallucination mitigation
• Own what you ship: prototype, deploy through Amazon AgentCore, monitor traces, and fix it when it breaks

Collaboration

• Partner with data engineers on Snowflake‑backed retrieval patterns (Cortex Analyst and Cortex Search Services)
• Contribute to refining our internal engineering patterns as the stack evolves

Skill sets – What you need to be successful

Required

• 3+ years of professional Python, with production experience building and operating services
• 1+ years of hands‑on work with LLMs in production: prompt/context engineering, tool/function calling, structured outputs, RAG
• Working knowledge of LangChain/LangGraph or a comparable framework like AgentCore Strands, CrewAI, or Semantic Kernel
• Experience with LLM observability tools: Amazon CloudWatch, LangSmith, Langfuse, MLflow, or OpenTelemetry
• Experience designing evaluation frameworks (MLFlow, DeepEval, LLM‑as‑judge, multi‑turn regression)
• Fluency with Git, Docker, and modern API frameworks
• Clear written communication and the judgment to know when something is ready to ship

A bachelor's degree is not required. Equivalent practical experience – bootcamps, self‑taught work, career changes, or non‑CS technical degrees – counts.

Strongly Preferred

• Hands‑on experience with Amazon Bedrock and/or AgentCore as a developer: runtime, gateways, memory, policy, guardrails, observability, awscli, evaluations
• Experience with Snowflake, Snowpark, or Snowflake Cortex
• Fluency in writing and reading SQL, as well as understanding semantic models
• Familiarity with multi‑agent patterns: supervisor/router, subagent/handoff, reflection, human‑in‑the‑loop
• A considered view on where agents should and shouldn’t act and comfort pushing back when “let’s add an agent” isn’t the right answer
• Experience in Loyalty, MarTech, AdTech, or a comparable data‑rich B2B domain

Equal Employment Opportunity

Employment at Kobie is based solely on an individual's merit and qualifications, which are directly related to professional competence. We do not discriminate against any teammate or applicant because of race, color, religion, gender, sexual orientation, gender identity/expression, national origin, disability, age, genetic information, veteran status, marital status, pregnancy, or any other characteristic protected by applicable law.

We are fiercely committed to fostering a workplace where teammates can bring their authentic selves to work every day. Our DEI initiatives, including various committees, ensure that principles of equity, diversity, and inclusion are deeply ingrained throughout Kobie. While our leadership team fully supports our policy of nondiscrimination and equal opportunity, it is the responsibility of all teammates to uphold these values.