Enterprise Architects Map the Whole Business, Solution Architects Map One Workflow

In an AI transformation, the enterprise architect owns the 30,000-foot view of how data, software, and AI plug into every department, while the AI solution architect owns one specific workflow: how it runs today, what the AI version looks like, what software it touches, which teams are involved, and what data it needs. Benjamin Fowler walks through both roles on screen using his Conducting AI templates, pointing at a Figma board that maps AI agents across an entire company and a course lesson that redesigns a single lead scoring workflow. He adds one caveat for smaller companies: these are separate jobs at large firms, but in a small team one person wears both hats.

The Enterprise Architect Owns the Company-Wide Map

The first artifact on screen is a FigJam board titled "Conducting AI Agent Template V6.0 Enterprise." It shows an organization chart of AI agents spanning Sales and Revenue, Marketing and Growth, Customer Success, Business Operations, People and Culture, Data and Analytics, Finance, Product Management, Engineering, and Design and UX. Each department has its own agent node with phased workflows beneath it, plus a legend of tools, integrations, and software assets. Fowler frames this as the enterprise architect's job: seeing where AI plugs into each department so someone can manage the change across the whole business rather than one team at a time.

Supporting Artifacts: Roadmap and Business Reference Model

The same board includes an "AI Development Roadmap," a color-coded swimlane chart with one row per agent across phases, and a "Business Reference Model" that breaks the company into 12 numbered functions, from Strategy and Leadership through Finance, Data and Analytics, and AI and Automation. A "High Level Functionality" panel splits the rollout into Phase 1, Phase 2, Phase 3, and an AI-Native end state. These are the enterprise architect's working documents: a current-state map of the business and a sequenced plan for where agents land.

The Solution Architect Redesigns One Workflow

The second artifact is a course lesson on a lead scoring agent. It documents the current workflow step by step: a trigger when an enriched lead hits the CRM, then rule-based scoring, SDR manual triage, score-versus-gut reconciliation, manager tier overrides, and queue placement, with the outcome that scoring "becomes a political artefact that everyone games and no one trusts." A "Why This Manual Workflow Exists" section explains the history, and "The AI-Native Solution" lays out the replacement: signal ingestion, predictive scoring from closed-won history, plain-English reasoning generation, confidence-gated routing, a human review queue, and live prioritisation. This is the solution architect's altitude, mapping how one workflow is currently done and then mapping the AI version with the software, teams, and data it requires.

The Documentation Goes Past the Happy Path

The lesson pages visible in the frames cover more than the workflow diagram. A "Legal and Compliance" section walks through GDPR legitimate interest, GDPR Article 22 safeguards for automated decisions, CCPA opt-out and deletion handling, vendor data sourcing, data residency, SOC 2 criteria, and the EU AI Act classification for B2B lead scoring. Another section addresses rep trust: make scoring criteria visible, let reps override with no friction, and use override patterns for system improvement rather than performance management. The lesson sidebar shows this is one of many, organized as Phase 1 Co-Pilots, Phase 2 Sales and Revenue Brain, and Phase 3 multi-agent systems, with named agents like Lead Enrichment and Routing, Prospect Research, Lead Scorer, and Sequence Content plus Reply-Intent.

The Pitch: 140 Use Cases for $99 a Month

Fowler closes with the offer. Conducting AI packages the enterprise architect frameworks and 140 documented AI agent use cases across six industries, priced at $99 a month as early-bird pricing he says will rise over time. The site page shown on screen positions it against McKinsey, Deloitte, and BCG transformation roadmaps that run $200K to $500K, and claims coverage of all 10 departments rather than just IT, quarterly updates, and board-ready ROI math and compliance notes for each workflow.

Key Takeaways

• Enterprise architects work at the whole-business level, mapping where data, software, and AI fit into every department so the change can be managed centrally.
• AI solution architects work at the single-workflow level: current process, AI redesign, software touched, teams involved, data required.
• Large companies staff these as separate roles. Small companies put both on one person.
• A complete solution design covers more than the agent workflow. The example lesson includes compliance (GDPR Article 22, CCPA, EU AI Act) and change management for the reps affected.
• Conducting AI sells these frameworks with 140 agent use cases across six industries at $99 a month, positioned as a substitute for six-figure consulting roadmaps.

Resources

• Conducting AI, Benjamin Fowler's platform with the agent templates, playbook, and use case documentation shown in the video