> Markdown version of [/jobs/ext/4649711-founding-ai-platform-engineer-agents-rag-fulltime-emea-latam-na](https://www.wearedevelopers.com/jobs/ext/4649711-founding-ai-platform-engineer-agents-rag-fulltime-emea-latam-na). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Founding AI Platform Engineer (Agents & RAG) - Fulltime - EMEA/LATAM/NA

- **Company:** Arc Full-time
- **Location:** United States (Remote available)
- **Experience:** Expert
- **Contract:** Contract
- **Skills:** Application Programming Interfaces (APIs), Artificial Intelligence, Audit Trail, Cloud Computing Security, Systems Integration, Retrieval-Augmented Generation, Large Language Models, Multi-Agent Systems, Generative AI, Data Layers, AI Platforms, Data Pipelines
- **Published:** November 9, 2026
- **Apply:** https://arc.dev/signup?userType=developer&publicRandomKey=py4titn2gv&to=https%3A%2F%2Farc.dev%2Fdashboard%2Fd%2Fverified-jobs%2Fpy4titn2gv%2Fapplication

## About the Role

* Hands-on experience building LLM agent systems: orchestration, tool use, multi-agent workflows.
 * Retrieval-augmented generation (RAG) and data-pipeline design in production.
 * Cloud security and identity & access management (IAM); audit logging and observability.
 * Integrations via APIs and MCP; experience with the Microsoft enterprise stack.
 * Ability to work independently as the sole engineer and own systems end to end., * Working knowledge of finance: CIMs, three-statement models, DCF/LBO, investment memos.
 * Experience in regulated or confidential-data environments (financial services, advisory).
 * Client-facing or forward-deployed engineering experience.

## Description

* Interim setup: Stand up one enterprise AI platform on a private cloud account; structure all work as missions with approval checklists, an evidence log and per-client folders.
 * Platform MVP: Build the first agents (mission manager, intake & data, analyst, evidence auditor), the evidence ledger and the approval console. First use case: an AI operations diagnostic for advisory clients.
 * Full platform (v1): Scale to a roster of ~12 agents under a manager agent; build a model gateway that selects the best model per agent, with review agents on a different model family; RAG from day one; tiered data layers (firm-wide, advisory, per-client encrypted vaults); hard-stop approval gates.
 * Client delivery: Run agent work in the firm's sandbox on copies of client data; support human-led deployment into client systems under time-limited, logged access; build read-only KPI feeds.
 * Investment automations: CIM analysis; deal pipeline management and outreach; three-statement analysis, DCF/LBO model building and investment memo drafting for screened deals.
 * Expansion: Keep the platform model-agnostic and take it from advisory to firm operations to portfolio companies.

Non-negotiable standards

 * Security and containment for client-system connections and financial data.
 * Information barriers enforced in code between business lines and data tiers.
 * Agents never receive write access to client systems; human review at every milestone.
 * Frontier models accessed through the gateway; no in-house model training.