Partner Technical Architect - Agentforce

Salesforce Salesforce · Enterprise · Chicago, New York - New York, Texas - Austin, California - San Francisco, IL

Salesforce is seeking a Partner Technical Architect to help SI partners build and ship production-grade Agentforce solutions. This role involves prototyping, building reference implementations, reviewing partner architectures, and coaching partners. The ideal candidate has strong engineering judgment, applied AI fluency, and builder instincts, with experience in LLMs, agents, RAG, tool use, and enterprise architecture.

What you'd actually do

  1. Prototype quickly to validate ideas, expose platform gaps, and unblock first-of-kind implementation risks.
  2. Build reference implementations, starter kits, demos, field kits, and evaluation harnesses that partners can fork, extend, and deploy.
  3. Convert field lessons into durable reusable assets: code, diagrams, checklists, decision trees, implementation notes, and architecture guidance.
  4. Review partner Agentforce architectures for agent design, retrieval strategy, tooling, guardrails, observability, cost, security, and production readiness.
  5. Coach Partner FDEs and architects through design clinics, office hours, portfolio reviews, and targeted technical walkthroughs.

Skills

Required

  • Applied AI fluency
  • Builder instincts
  • Ability to ramp quickly into the Salesforce platform
  • Experience building with LLMs, agents, RAG, tool use, workflow orchestration, or automation systems
  • Ability to reason about multi-agent, human-in-the-loop, and enterprise integration patterns
  • Comfort debugging agent behavior: traces, logs, failure modes, tool calls, bad context, bad data, latency, and user experience gaps
  • Practical understanding of prompt design, retrieval quality, evaluation, hallucination risk, guardrails, and cost/performance trade-offs
  • Strong systems thinking across APIs, data flows, identity, security, integration, observability, and operational handoff
  • Strong product sense

Nice to have

  • Salesforce, Agentforce, Data Cloud, Apex, Flow, LWC, MuleSoft, or multi-cloud architecture experience
  • Salesforce certifications
  • Experience creating maturity models, readiness checks, production-readiness scorecards, or technical review frameworks
  • Familiarity with consumption economics, agent cost modeling, TCO, or outcome-based delivery models
  • Experience working with system integrators, partner delivery teams, or complex enterprise implementation programs
  • Published technical work: open-source projects, internal platform docs, talks, blog posts, or reusable architecture patterns

What the JD emphasized

  • production-grade Agentforce solutions at scale
  • build, debug, explain trade-offs
  • applied AI fluency
  • agent design
  • retrieval strategy
  • tooling
  • guardrails
  • observability
  • production readiness
  • shipped or seriously prototyped real software, agents, automations, internal tools, or technical demos
  • move fast without losing production instincts around safety, reliability, observability, and maintainability
  • extract patterns from one-off work and turn them into reusable assets for others
  • building with LLMs, agents, RAG, tool use, workflow orchestration, or automation systems
  • reason about multi-agent, human-in-the-loop, and enterprise integration patterns
  • debugging agent behavior
  • prompt design
  • retrieval quality
  • evaluation
  • hallucination risk
  • guardrails
  • cost/performance trade-offs
  • regulated or complex enterprise environments

Other signals

  • Build and ship production-grade Agentforce solutions at scale
  • Turn ambiguous enterprise AI problems into working solutions
  • Package what they learned so many partner teams learn from it