Northstar
AI recommendations & content activation.

Adaptive Content Serving (ACS) already supported nine recommendations below a standard article. Expansion required consistent placement across varied templates, dependable content metadata and commercial serving rules.

My product work connected those requirements with the engineering, data, editorial and business teams responsible for operating the platform during the June 25–August 1, 2025 contract.

I audited templates across Northstar Meetings Group, Business Travel News, TravelPulse, Meetings & Conventions, and Successful Meetings. My recommendations for conditional widget placement, duplicate-module prevention, and clearer sponsored-content labeling were adopted.

I defined a proposed Container → Slot → Widget → Content contract: container width, eligible widget types and maximum slots; horizontal or vertical slot combinations with padding; and defined content types for each widget.

The expansion requirements also covered sponsored spotlights, newsletter calls to action, events, webinars and richer content stacks, including metadata, eligibility, timing, urgency and placement controls.

The existing daily pipeline ran from Ektron through LeverageLabs and BigQuery to Template Builder. Ektron held content and metadata; LeverageLabs processed content and supplied the recommendation table; BigQuery contained the output; internal developers connected that table to Template Builder to render the module.

I defined requirements across ingestion, scoring inputs, cohort assignment and eligible template placement, connecting Navigator audience inputs and recommendation tables with CMS rendering rules.

I coordinated checks for valid URLs, consistent content classification, scoring completeness and ingestion confirmation or delays. Inconsistent article definitions and sponsored-content classification required content and data cleanup across Ektron and NetSuite.

The proposed Marketing Hub input model covered headlines, descriptions, sources, tags, dates, URLs or assets, content and promotion types, intent and audience. Supporting field groups addressed campaigns, timeliness and content-specific requirements for video, webinars, events, newsletters and research.

I translated dependencies into SOPs, integration diagrams and troubleshooting guidance covering metadata completeness, ingestion, classification, approvals and campaign timing. Checks for missing content, incorrect types, delays and publishing failures helped teams identify the failing step and coordinate a correction.

I led requirements and prototype work for a Marketing/Activation Hub within Ektron workflows. The business-user flow moved from brand and content-type selection through metadata and assets, audience and duration, campaign placements and submission for ingestion.

The prototype showed promotion inventory and campaign configuration across web widgets, social and email. Proposed generation workflows retained human editing and approval before publication; persona guidance and related AI Planner use cases remained requirements and prototype planning.

Adopted recommendations for conditional placement, duplicate-module prevention, and sponsored-content labeling were the concrete product result. I also delivered technical requirements, validation criteria, expansion specifications, SOPs, integration diagrams and troubleshooting guidance, alongside the Marketing Hub activation prototype.

The existing MVP snapshot recorded 51,346 page serves and 288 clicks, providing a baseline view of reader engagement for expansion planning.

The lesson was to connect recommendation data with the rules and operating responsibilities that determine what readers actually see. Clear handoffs, placement controls and review steps made the expansion requirements actionable across teams.