Salesforce (AI) Platform Engineer

Job description

**Position Overview:** Randstad Digital is building an agentic framework a Salesforce-native, multi-agent AI platform: an orchestrator “brain” coordinating Agentforce specialists, LLM engines and deterministic Apex skills across channels (in-org voice assistant, Google Chat, mobile). We are seeking a Salesforce AI Platform Engineer with 4+ years of experience to help build the platform itself — not configuration, not release management — working in an AI-first model, pairing daily with Claude-based engineering agents alongside our IT Director and engineers. Key Responsibilities: - **Platform Engineering (Apex):** Design and extend orchestration layers — planner loops, deterministic intent routing, circuit breakers, audit trails, session continuity — with rigorous callout/DML ordering and a test-first discipline. - **LLM Integration:** Build against Anthropic Claude and Google Gemini APIs and more — tool-use loops, system-prompt engineering, structured outputs and fallback chains; own the quality and latency of AI answers. - **Agentforce:** Create and maintain agents via the Agent API and GenAI metadata (planner bundles, topics, plugins); navigate platform quirks and compile behaviours. - **Channel Engineering:** Evolve channel agents — Google Chat app (Cards v2, dialogs, slash commands, service-account OAuth), Cloudflare Worker relays, and future Teams/Slack surfaces. - **Front-End (LWC):** Build polished, branded assistant UIs — utility-bar voice assistant, orchestrator console dashboards — with Lightning Web Security-aware patterns. - **AI-Assisted Delivery:** Drive Claude Code engineering sessions, review AI-authored code critically, and keep specs and handoff documents current so multiple AI-assisted engineers work concurrently without collisions. - **Integration Plumbing:** Named/External Credentials, External Client Apps, OAuth 2.0 flows (client credentials, JWT bearer) and inbound REST webhooks with JWT verification. Technical Skills & Qualifications: **Required:** - 4+ years of Salesforce platform development: advanced Apex (async patterns, callouts, dynamic SOQL, USER\\_MODE security), LWC, Metadata API and SFDX-based deployments. - Hands-on LLM API integration experience (Anthropic, OpenAI or Google Gemini) — tool/function calling, prompt engineering, multi-turn state management. - Solid JavaScript/TypeScript and Node.js; comfort with serverless edges (Cloudflare Workers or similar). - Demonstrated fluency working WITH AI coding assistants (Claude Code, Copilot, Cursor) as a force multiplier — you review, verify and direct; you never paste blindly. - OAuth 2.0 depth: client credentials, JWT bearer, service accounts, token lifecycle. - Strong written communication — specs, handoffs and decision logs are how this team scales. - Agile self-starter comfortable with high autonomy and rapid shipping from sandbox to UAT to production. **Nice to have:** - **Certifications:** Salesforce Platform Developer II • Salesforce AI Specialist / Agentforce Specialist • Salesforce JavaScript Developer I. - Google Workspace app development (Chat apps, Cards v2, service accounts). - Voice interfaces: TTS/STT integration (ElevenLabs, Web Speech API). - React Native (mobile companion app). - Security frameworks (OWASP SAMM, ISO 27001)