How an AI-native SMO carried a Supply Chain Data and Risk Management Company into healthcare.

The work would have demanded a full PMO staffed by one to three full-time resources per workstream. An AI-native Strategic Management Office stood up instead, with humans serving as the quality check and external customer-facing support.

How the 3Pillars team mapped a healthcare entry from the outside in.

The company knew logistics. It did not know healthcare procurement cycles, state emergency-preparedness funding, surgical-vendor sales motions, or the cross-functional choreography of a hospital pilot. The diagnostic was two jobs in one: learn the commercial landscape, and learn the company’s own workflows well enough to retrofit an AI-native operating layer underneath them.

• A documented workflow map for both product lines covering external opportunity sensing and internal coordination
• A prioritized list of RFP sources, grant programs, conferences, and partnership targets
• A sales-plan overlay showing exactly where the SMO will feed sales ops reps
• An AI readiness baseline and a six-workflow build plan
• A human-in-the-loop design that defines what humans approve, what humans teach, and what runs autonomously

How the SMO came online across two product lines.

The build was deliberately split. Three workflows pointed outward at the market: RFPs and grants, conferences, and market-entry / partnership opportunities. Three workflows pointed inward at the company: a silent meeting participant, cross-functional coordination and risk surfacing, and board-and-customer deck generation. Each stood up with a defined human-in-the-loop checkpoint so the organization could trust the output and teach the system where it was wrong.

• External opportunity sensing. RFP and state-grant monitors, conference and attendee trackers, and a market-entry / partnership agent feeding both product lines.
• Internal coordination layer. Silent meeting participant capturing notes, owners, and follow-up routing. Cross-functional project plan and risk surfacing. Board and customer deck generation.
• Sales ops enablement. Routing layer that takes external signals and matches them to the right sales ops rep with context, suggested outreach, and tracking back into the project plan.
• Quality and teaching loop. A lightweight review workflow where humans correct the system, and corrections feed back into prompts, skills, and reference data so the SMO gets sharper each cycle.

• SMO command surface. A shared workspace built on Notion, Granola, and Claude where the PMO lead, sales ops, and executive sponsor see status across both product lines in one place.
• Document repositories. Instrumented for RFPs, grants, conference intel, partnership notes, project plans, board decks, and customer-facing materials.
• Human-in-the-loop review cadence. Defined approvers per workflow. Owners can correct or reassign within a defined window, and corrections train the system.

Six workflows, one silent operating layer.

Three outward, three inward. Each with clear inputs, clear outputs, and a defined human-in-the-loop checkpoint. Nothing goes out the door without a person signing off.

Four roles, one operating layer underneath them.

Pick a persona in the left nav to see how the SMO changes that seat.

What the company was left with.

The SMO is a persistent operating layer that runs daily, weekly, and quarterly across both product lines, with humans serving as the quality-check and teaching layer.

• More product lines. Add a third or fourth product line by extending ICP definitions and source lists. The six workflows generalize.
• Deeper account coordination. Layer customer-success workflows underneath the existing SMO so post-sale execution sits in the same operating fabric.
• Channel and GPO motions. Add GPO and channel-partner workflows as the company moves from direct-sales pilots to scaled distribution.

• Deeper data-agent capabilities. Move from device-offline analysis and reporting into predictive maintenance, utilization analytics, and proactive client alerts.
• Customer-facing reporting and dashboard automation. Extend the documentation agent into automated client-facing reports and dashboards, reducing the manual reporting load on customer success and product teams.
• Closed-loop sales intelligence. Feed outcome data from won and lost RFPs back into the opportunity-scoring workflow so the SMO gets sharper at fit prediction over time.
• AI-native security and provisioning workflows. As the IT security and hardware provisioning teams gain comfort with the SMO, fold their internal processes into the same operating layer.