We're building AI agents across our Salesforce applications to automate case handling, document review, and recommendation workflows. Our first deployment automates supply chain invoice exception processing; additional agents will follow across other Salesforce applications we've built and deployed. We're looking for someone to design, build, and iterate on these agents end to end.

What you'll do:

- Build agents in Salesforce Agentforce — topics, actions, and grounding against Salesforce and external data.
- Integrate agents with our ERP and document repositories so they validate against records, not just user claims.
- Design the human-in-the-loop review experience where business owners approve or modify agent recommendations.
- Establish evaluation sets, monitor accuracy, and tune automation posture based on measured performance.
- Work with business stakeholders to translate operational rules into agent behavior.

Required

- Hands-on Salesforce development: Apex, Flow,



Lightning Web Components, and the Salesforce data model
- Salesforce integration experience — REST/SOAP APIs, middleware, external data access
- Production experience building LLM-based agents: prompt design, grounding, tool calling
- Retrieval over unstructured documents (email, PDFs, attachments)
- A real approach to evaluating AI output quality, not just shipping it
- High level of English B2+

Preferred

- Agentforce or Einstein AI, including Data Cloud grounding
- Healthcare or other regulated-industry experience — audit trails, approval controls
- Salesforce certifications: Platform Developer I/II, Agentforce Specialist, AI Associate

What we're looking for Judgment about when not to automate. We need someone who designs for human approval when needed (avoid risk when required).

📌 AI Agent Developer — Salesforce (Colombia)
🏢 Recruitment Bpo Col
📍 Colombia