Salesforce Agentforce Architect

NTT DATA Services•Atlanta, GA

About The Position

The Salesforce Agentforce Architect owns end-to-end solution architecture across Agentforce, Salesforce CPQ, and Revenue Cloud Advanced. The role architects agent-assisted and agentic workflows across opportunity management, configuration, pricing, quoting, contracting, ordering, and revenue lifecycle processes. The architect designs agent orchestration, integrations, context and retrieval strategies, secure action patterns, evaluation frameworks, and observability. The role assesses model and platform trade-offs, designs resilient transaction patterns for revenue processes, and advises on consumption economics and cost control. It also leads architecture reviews and technical governance across global delivery teams and mentors technical architects, developers, and Agentforce specialists.

Requirements

Expertise in Salesforce Agentforce and associated agentic architecture concepts.
Expertise in Salesforce CPQ and Revenue Cloud Advanced / Revenue Lifecycle Management.
Expertise in Salesforce Sales Cloud and core platform architecture.
Knowledge of agent orchestration and tool/function calling.
Knowledge of multi-agent architectures, delegation, and coordination patterns.
Knowledge of agent state, short- and long-term memory, and session/context management.
Knowledge of MCP and agent-to-agent interoperability concepts and patterns.
Knowledge of RAG, semantic/vector search, embeddings, chunking, and retrieval optimization.
Knowledge of context engineering beyond basic prompt engineering.
Knowledge of evaluation frameworks and AI observability.
Knowledge of AI safety, grounding, guardrails, and least-privilege action design.
Knowledge of LLM model selection and architectural trade-offs.
Knowledge of classic NLP concepts, including intent classification, entity extraction, semantic similarity, and ranking.
Knowledge of Apex, Lightning Platform, Flow, APIs, and enterprise integration patterns.
Strong understanding of asynchronous processing, transaction boundaries, resiliency, retries, and idempotency.
Knowledge of security architecture, including sharing, CRUD/FLS, permission models, and data protection.

Nice To Haves

15+ years of overall technology experience and 10+ years in Salesforce CPQ / Revenue Cloud architecture.
Proven experience architecting enterprise CPQ / Quote-to-Cash implementations.
Salesforce Application, System, or Technical Architect credentials.
Agentforce and Revenue Cloud-related certifications or demonstrable implementation experience.
Experience with MuleSoft and large enterprise integration landscapes.
Ability to communicate architectural decisions to engineering teams and senior business stakeholders.

Responsibilities

Own end-to-end solution architecture across Agentforce, Salesforce CPQ, and Revenue Cloud Advanced.
Architect agent-assisted and agentic workflows across opportunity, configuration, pricing, quoting, contracting, ordering, and revenue lifecycle processes.
Design agent orchestration patterns, including tool/function calling, workflow delegation, multi-agent patterns, state management, and memory.
Architect integrations of Agentforce with Salesforce automation, Apex, Flow, APIs, MuleSoft, and external enterprise services.
Determine appropriate use of LLMs, deterministic logic, workflow engines, and traditional automation.
Design context engineering and retrieval strategies, including structured and unstructured grounding, retrieval depth, metadata filtering, and relevance optimization.
Establish secure agent action patterns using least-privilege access, guardrails, validation, approval checkpoints, and human-in-the-loop mechanisms.
Define AI evaluation frameworks covering answer quality, grounding, hallucination, task completion, tool selection, latency, and business outcomes.
Establish observability across agent reasoning paths, actions, integration failures, token/credit consumption, and operational performance.
Assess model and platform trade-offs across quality, latency, context window, cost, privacy, and enterprise constraints.
Architect resilient action layers and transactional integration patterns for revenue processes where consistency and recoverability are critical.
Advise programs on consumption-based economics, AI credit/token optimization, and architectural approaches to control run-time cost.
Lead architecture reviews, design authorities, and technical governance across global delivery teams.
Mentor technical architects, developers, and Agentforce specialists.