AI Engineer – Internal Agents & Workflows

This is a remote position. Location: Permanent Remote Engagement Type: Fixed-Term Contract (6 Months) – Dedicated Role (No Freelance/Ad-hoc) Start Date: Immediate About the Project Our Client is building custom internal AI agents and workflow automations to boost productivity across our Support, Sales, Customer Success, Recruiting, and Engineering teams .

Our goal is to automate repetitive work—such as triage, summaries, follow-ups, and reporting—while enabling "copilot" experiences that help our teams execute faster with consistent quality . This is a hands-on build role where you will integrate with our tools (Slack, Jira, Zoom, Email, Docs, CRM), implement agent workflows, and ship working automations quickly .

What You’ll Build (Key Use Cases) You will focus on high-impact MVPs that can be deployed and improved in short iteration cycles . Examples include: Support & Engineering: An AI triage bot to classify bugs vs. change requests, ask required questions, and auto-create tickets with the correct fields . Sales & Customer Success: A Sales copilot that provides real-time answers grounded in product specs , and automations to generate follow-up emails, route them for approval, and send them .

Recruiting: An interview evaluation agent that ingests transcripts to apply role rubrics and generate structured feedback . Internal Knowledge: Self-serve assistants and chatbots within Slack .

Key Responsibilities

Design and implement workflows involving tool calling, multi-step reasoning, and structured outputs . System Integration: Build robust integrations and connectors with common internal systems such as Slack, Jira, email, calendar, CRM, and data warehouses . RAG Implementation: Implement RAG (Retrieval-Augmented Generation) and knowledge systems for product and internal docs, ensuring citations and links to sources are included .

Security & Access: Build basic access control patterns, including SSO/AD group-aware retrieval and role-based response filtering . Production & Guardrails: Ship MVPs fast, iterate based on user feedback, and add necessary guardrails such as evals, logging, and retry/timeout behaviors . Documentation: Document workflows and provide simple runbooks to facilitate internal adoption .

Required Skills

& Qualifications Software Fundamentals: Strong software engineering fundamentals with a focus on clean, testable code and a production mindset . LLM & Agent Experience: Proven experience building with LLMs and agents, specifically regarding prompting, structured output (JSON schemas), and tool/function calling . Orchestration: Experience with multi-step orchestration (agent loops, state machines, graphs) .

Frameworks: Hands-on experience with at least one major framework such as Claude Agent SDK / MCP, Lang. Graph, or OpenAI Assistants . API Integration: Strong experience with APIs and integrations (OAuth/SSO patterns are a plus) . Workflow Engines: Familiarity with workflow engines like Temporal, Dagster, Airflow, or Prefect .

Retrieval Systems: Experience with retrieval systems such as Llama. Index, Lang. Chain RAG, Elasticsearch, pgvector, or Pinecone . Tech Stack We are flexible but expect a mix of the following: Languages: Python and/or TypeScript . Agent Frameworks: Claude Agent SDK and/or Lang. Graph . RAG: Llama. Index or existing vector retrieval stacks .

Integrations: Slack, Jira, Email, Call Recording, and CRM APIs . Success Metrics (First 2–4 Weeks) Deliver 1–2 production-ready MVP automations (e.g., interview evaluator or support triage bot) . Establish basic observability including logs, failure handling, and simple admin runbooks . Create a clear plan for scaling to more use cases with reusable connectors .