Principal Systems Architect

As Principal Systems Architect for GTM & Business Systems at Engine, you'll own the end-to-end architecture of the platforms that power how we go to market: Salesforce (Sales Cloud, Service Cloud, Data Cloud), Tray.io, Outreach, Slack, ZoomInfo, AWS Connect, that let these systems talk to each other, to our product, and to our engineering stack.

You're not maintaining someone else's setup. You're designing the blueprint, shipping production-grade automation, and setting the technical standard for how GTM systems and the multi-agent architecture that operates on them run at scale.

This is a senior individual contributor role. You architect, you build, and you ship. You partner directly with Sales, Marketing, RevOps, Product, and Engineering to translate ambiguous business problems into durable, metadata-driven systems that scale faster than headcount with Claude-powered agents and agent-to-agent orchestration as a first-class output, not a bolt-on.

What You'll Own

Salesforce Platform Architecture. Own the end-to-end design of Engine's Salesforce stack — Sales Cloud, Service Cloud, Data Cloud, and Experience Cloud. Architect declarative-first, metadata-driven solutions (Flow, Permission Sets, TriggerAction patterns) that minimize technical debt and maximize admin velocity. Set the standard for governor limits, bulkification, sharing, and security — including the API and permission surfaces that MCP servers expose to Claude-powered agents.

Tray.io Orchestration & Integration Layer. Architect the middleware fabric that connects Salesforce to Outreach, ZoomInfo, Snowflake, Customer.io, AWS Connect, NetSuite, and dozens of GTM tools — and increasingly, to the agents that operate across them. Own workflow design, error handling, idempotency, retries, and observability. Make the call on when to use Tray vs. native Salesforce automation vs. Apex vs. event-driven patterns through Kafka vs. Claude-orchestrated agent flows.

Sales Engagement & Outreach Architecture. Own the technical architecture of Outreach — sequences, triggers, mappings, ZoomInfo enrichment, and the bi-directional sync with Salesforce. Design the signal layer and Claude-driven prioritization that powers AE/AM workflows, lead routing, and the ranked actions that show up in the Cockpit every morning.

Engineering & Product Interface. Serve as the technical translator between Business Systems and Engineering. Partner with Product and Engineering on event streams (Kafka, webhooks), shared data contracts, Snowflake pipelines, MCP endpoints, and the customer-facing surfaces that touch Salesforce (Help Center, EVA chat/voice, Atlas). Own the integration patterns that let GTM systems — and the agents that operate on them — consume product data without becoming a bottleneck.

Slack as the GTM Control Plane. Architect Slack as the operational surface for GTM — automated pulse reports, deal alerts, approval workflows, case routing, incident response, and Claude-powered agents that GTM teams can query in-channel and hand off to specialized agents downstream. Build Slack-native experiences (Block Kit, slash commands, modals, agent invocations) that move work out of inboxes and into the channels where decisions actually happen.

Sprint Execution & Communication. Manage your own backlog, prioritize with GTM and RevOps leadership, and ship on a two-week cadence. Communicate async-first. Surface blockers and design trade-offs before they become fires.

What You'll Bring

Technical Depth

• 8+ years architecting Salesforce at scale — Sales Cloud, Service Cloud, Flow, Apex, LWC, Experience Cloud, and Data Cloud. You've designed for governor limits, multi-cloud sharing, and complex sObject models, not just configured out of the box.
• Deep, hands-on experience with Tray.io (or equivalent: Workato, Mulesoft, n8n) — workflow design patterns, error handling, callable workflows, and high-volume orchestration across millions of records.
• Production experience with Outreach architecture — triggers, sequences, plugins, ZoomInfo enrichment, and Salesforce sync.
• Strong working knowledge of Slack as a platform — Block Kit, slash commands, event subscriptions, and Slack-native workflow design.
• Hands-on experience building on Claude (or an equivalent frontier LLM) in production — model selection, prompt design, tool use, evals, prompt caching, structured outputs, and the cost/latency tradeoffs that come with running agents at scale.
• Production experience with MCP — you've built MCP servers, not just consumed them. You understand tool contracts, auth scopes, permissioning, and the operational surface area of an MCP endpoint that touches real customer data.
• Real reps with multi-agent / agent-to-agent orchestration — capability discovery, task delegation, handoff protocols, shared state, and the patterns that keep multi-agent systems debuggable when something goes sideways.
• Proven experience integrating GTM systems with data warehouses (Snowflake), event streams (Kafka, webhooks), and product systems — REST/SOAP APIs, event-driven architecture, and idempotent design.

Systems Thinking & Integration

• You think in data models, event flows, and contracts — not screens. You design for downstream reporting, audit requirements, and cross-system consistency, including the contracts agents rely on.
• Comfortable owning the full integration lifecycle: requirements, API review, sandbox, production deployment, observability, and post-launch optimization.
• Strong instincts on clicks vs. code, declarative vs. programmatic, Tray vs. Apex vs. Kafka vs. agent — and the ability to defend each call with business and technical reasoning.

Engineering Collaboration

• You speak the language of Product and Engineering. You read PRDs, review code, comment in GitHub PRs, and hold your own in architecture reviews.
• Hands-on with version control, sandbox strategy, CI/CD for Salesforce (sfdx, scratch orgs, metadata deployments), and modern DevOps patterns.
• Familiar with event-driven patterns, message queues, MCP servers, and the realities of distributed systems — eventual consistency, retries, dead-letter queues, and the difference between at-least-once and exactly-once delivery.

Speed & Execution

• Action-biased. You ship, iterate, and don't wait for perfect specs.
• AI-driven builder — fluent with Claude Code, Cursor, MCP servers, and the Anthropic API. You use AI to accelerate time-to-production on integrations, reviews, and documentation, and you have strong opinions on where it does and doesn't belong.

Communication

• Tight, clear async writer — design decisions, blockers, and risk flags without being prompted.
• Shows up to syncs with recommendations, not status.
• Escalates early and with proposed solutions.
• Translates technical architecture into business impact language for Sales, RevOps, Marketing, and CX leadership.

Bonus

• Shipped multi-agent systems or production MCP servers at meaningful scale.
• Contributed to open-source MCP servers, A2A implementations, or agent frameworks.
• Background building AI SDR, lead scoring, or account prioritization systems.
• Experience architecting call intelligence and conversational AI platforms (Attention, Gong, AWS Connect, Twilio).
• Previous experience as the bridge between Business Systems and Engineering at a high-growth, venture-backed company.
• Salesforce certifications (Platform Developer II, Application Architect, System Architect, or CTA-track).

#LI-MH1

Applications for this role will be accepted through October18th, 2026 or until the role is filled. We encourage you to apply early, as we may begin reviewing applications before the deadline.