Applied AI Engineer
Role summary
We are seeking an Applied AI/LLM Engineer to design, build, and deploy production-grade LLM-powered agents and applications. This role involves significant coding to orchestrate multi-step workflows, integrate APIs, and ensure agent predictability. You will work closely with the Data Science team, translating their LLM strategy and guardrails into robust systems. Responsibilities include implementing RAG pipelines, managing the full lifecycle of agent systems, building evaluation and observability infrastructure, and collaborating on prompt engineering and model tuning. Experience with cloud environments (GCP/Vertex AI, AWS Bedrock, Azure OpenAI) and a strong software engineering background are essential for this builder-focused position.
### Who you are
- 3+ years of software engineering experience with strong proficiency in Python Hands-on experience building applications powered by large language models (Claude, GPT, Gemini)
- Familiarity with agent frameworks and orchestration patterns (LangChain, LangGraph, CrewAI, Vertex AI Agent Builder, or custom orchestration)
- Experience implementing function calling, tool use, and multi-step agent workflows Solid understanding of RAG architectures, embedding models, and vector databases (Pinecone, Weaviate, pgvector, Vertex AI Vector Search)
- Comfort working within defined guardrails and model configurations - you don't need to pick the model, but you need to know how to get the most out of it
- Experience with API design, microservices, and deploying services in cloud environments Strong debugging and problem-solving instincts. Agent systems fail in non-obvious ways and you enjoy chasing down why
- Strong communication skills; ability to work across Data Science, Product, and Engineering teams
- Experience with evaluation frameworks for LLM-based systems (custom evals, RAGAS, LangSmith, Braintrust), is nice to have
- Familiarity with MLOps tooling and CI/CD for ML systems and experience with streaming responses, async architectures, and real-time agent interactions, are nice to have
- Background in building multi-agent systems with routing, delegation, and coordination patterns, is preferred
- Exposure to Google Cloud Platform/Vertex AI ecosystem Contributions to open-source AI/ML projects, is preferred
### What the job involves
- We're looking for an Applied AI/LLM Engineer to design, build, and ship LLM-powered agents and applications
- You'll work closely with our Data Science team, who defines the LLM strategy, guardrails, evaluation criteria, and model selection
- Your job is to take that foundation and turn it into reliable, production-grade agent systems that solve real business problems
- This is a builder role, not a research role
- You'll spend most of your time writing code — wiring up tools, orchestrating multi-step workflows, integrating APIs, and making sure agents behave predictably in the wild
- You should be deeply comfortable working with large language models in production and operating at the intersection of software engineering and applied AI
- Build agentic applications and workflows using the LLM frameworks, models, and guardrails provided by the Data Science team Design and implement tool integrations, function-calling patterns, and orchestration logic that allow agents to take actions across internal systems and external APIs
- Translate agent specifications and prompt strategies (authored by Data Science) into robust, deployable services Implement RAG pipelines, including vector store integration, chunking strategies, and retrieval optimization
- Own the full lifecycle of agent systems from prototype through production, including testing, monitoring, logging, and iteration
- Build evaluation and observability infrastructure so the team can measure agent quality, latency, cost, and safety in production
- Collaborate with Data Science on prompt engineering, model behavior tuning, and guardrail enforcement. Implementing their specifications into the runtime layer Work with platform and infrastructure teams to deploy, scale, and maintain agent services in cloud environments (GCP/Vertex AI, AWS Bedrock, Azure OpenAI)
- Contribute to internal tooling, SDKs, and shared libraries that accelerate agent development across the organization
### Benefits
- Medical (Free medical plan option for employees and affordable plan options for families)
- Health Savings Account (HSA) (Employer HSA contribution)
- Flexible Spending Accounts (FSA)
- Dental
- Voluntary Vision
- Life Insurance
(Company paid)
- Accidental Death and Dismemberment (AD&D) Insurance (Company paid)
- Voluntary Life and Accidental Death and Dismemberment (AD&D) Insurance
- Short-Term Disability (STD) (Company paid)
- Long-Term Disability (LTD) (Company paid)
- Employee Assistance Program (EAP)
(Company paid)
- Accident Insurance
- Critical Illness Insurance
- Hospital Indemnity Insurance
- Pet Insurance
Sample PagerDuty interview questions
- 1
Outline the architecture for a distributed ML system that ensures reproducibility and version control of models and data.
system designmedium - 2
Iterator over a Binary Search Tree Implement an iterator over a binary search tree. Input: root = [7,3,15,null,null,9,20], calls: next(), hasNext(), next() Output: 3, TRUE, 7 Explanation: The iterator yields the smallest value (3), confirms more nodes exist, then properly yields the next in-order value (7).
codingmedium - 3
Minimum Path Sum Find the minimum path sum from top-left to bottom-right in a grid. Input: grid = [[1,2,3],[4,5,6]] Output: 12 Explanation: The optimal path moves along 1 -> 2 -> 3 -> 6, resulting in a minimum accumulated sum of 12.
codingmedium - 4
Continuous Subarrays Sum Equals K Find the total number of continuous subarrays whose sum equals K. Input: nums = [1,2,3], k = 3 Output: 2 Explanation: Both the contiguous subarray [1,2] and the single-element subarray [3] sum perfectly to the target of 3.
codingmedium - 5
Aggressive Cows Maximize the minimum distance between aggressive cows in stalls. Input: stalls = [0,4,3,7,10,9], cows = 3 Output: 4 Explanation: Placing the cows at positions 0, 4, and 10 yields a maximum possible minimum distance of 4 between any two cows.
codingmedium
Sign up for a personalized interview prep pack tailored to this role.
Similar roles
Applied AI EngineerBarhead, an Akkodis company · Boston, Massachusetts, United States · Remote- Applied AI EngineerSaunders Scott · New York, New York, United States · Hybrid
Staff Applied AI EngineerVantor · Westminster, Colorado, United States · Onsite- Applied AI EngineerCatalyst Labs · Valle Vista, California, United States · Onsite
- Applied AI EngineerLawrence Harvey · New York, United States · Onsite