Two years building the platform strategy, delivery infrastructure, and AI workflows that let a SaaS ecommerce company actually use its own technology.
360training is an online compliance training company — a regulated SaaS business selling courses to B2B and B2C customers across industries including real estate, food safety, OSHA, DOT, and HIPAA. The web platform was a multi-site ecommerce environment running on Drupal and WebSphere, with an LMS underneath, disconnected data systems, and regulatory constraints on content that most web teams never have to think about.
The organization had real ambitions: improve conversion, modernize the platform architecture, grow B2B, launch acquired sites faster. The gap was between those ambitions and the operational infrastructure needed to execute on them — no delivery templates, no requirements standards, no AI-assisted workflows, and no clear path from the current platform to the decoupled architecture the business actually needed.
I was hired as Web Marketing Manager. The scope of what I actually did was closer to web program director.
I owned web delivery for the B2B and B2C ecommerce properties, partnering with SEO, content, design, engineering, and product management teams to improve acquisition, UX, and conversion outcomes. This included program-managing major redesigns, platform migrations, and post-acquisition site integrations — from requirements through UAT, launch, and stabilization.
I also regularly stepped into the space between business goals and technical execution that product management wasn't filling: translating vague feature requests into buildable IT specifications, pushing back on requirements that lacked UX or acceptance criteria, and surfacing the questions that needed answers before work entered development. This was the informal consulting layer that made the formal delivery work actually land.
The business was operating a tightly coupled CMS-ecommerce environment with product data locked inside WebSphere and content locked inside Drupal — neither connected to the other, and neither capable of serving the multi-site, multi-acquisition growth model the company was pursuing. The architecture was a ceiling.
I led the discovery and strategy work for a headless CMS transition. That meant understanding the full technical stack, reasoning through the business case for decoupling, and building the executive narrative — separately for the CTO and the CMO, because those are different conversations requiring different framing. The CTO needed the architecture rationale. The CMO needed the business outcomes: content hub efficiency, acquisition migration speed, COPE strategy, and the conversion improvements that come from owning product data properly.
Both approved it in separate review meetings. I was beginning to educate the rest of the team on the transition when my time at the company ended.
The architecture problem
Decoupled CMS + content hub
Tightly coupled Drupal and WebSphere with disconnected product data. The decoupled architecture strategy included a content hub model, COPE implementation, and a migration path that would make future acquisitions dramatically faster to integrate.
The executive case
CTO and CMO aligned separately
Built two distinct narratives for two distinct audiences. Architecture rationale and risk framing for the CTO. Business outcomes — acquisition speed, content efficiency, conversion lift, structured data — for the CMO. Both approved. Team education underway when the engagement ended.
The platform strategy work was happening on top of a delivery environment that needed its own stabilization. The team was doing everything manually and inconsistently — high-volume content, large releases, regulatory content changes, and post-acquisition integrations all running through ad-hoc processes that worked until they didn't.
I built the delivery infrastructure from scratch: intake processes with SLAs, estimation frameworks, UAT templates, launch checklists, and retrospectives. When the regulatory team was flooding the queue with "urgent" requests, I designed an Asana intake workflow with required fields and emergency routing, collaborated with Regulatory to define a 48-hour turnaround SLA for routine requests, and integrated them into the review and sign-off process. The result was shared accountability instead of shared frustration.
Redesigns and migrations
Program-managed major website redesigns and platform migrations across Drupal and WebSphere — requirements, project plans, acceptance criteria, UAT, and launch. Kept complex cross-team programs on track in a high-volume environment.
Regulatory intake workflow
Designed a structured Asana intake process with required fields, emergency routing, and a 48-hour SLA for routine regulatory content changes. Reduced cross-functional friction, rework, and the kind of timeline disruption that "urgent" without process creates.
B2B ecommerce improvements
Separated support requests from leads in the B2B funnel, aligned B2B content to sales goals, and improved UX and conversion performance across category and product pages.
Requirements translation
Regularly translated vague business feature requests into buildable IT specifications — surfacing missing UX requirements, acceptance criteria, and performance standards before work entered development. The PM-adjacent work that kept builds from going sideways.
The AI work at 360training is the story people ask about, so I want to tell it accurately. I didn't introduce AI because it was 2022 and everyone was talking about it. I introduced it because specific problems — high-volume recurring content, navigation architecture decisions, operational documentation — were exactly the kinds of problems AI is genuinely good at accelerating. The constraint I kept throughout was simple: AI touches the draft, humans own the judgment.
Promotional copy
Recurring regulated-industry promo copy reduced to approximately 10% of previous drafting time for certain content types, while maintaining brand voice, compliance accuracy, and human editorial review on everything that went live.
Navigation and SEO strategy
Used AI to reason through navigation IA, SEO implications of category restructuring, and E-E-A-T content strategy — accelerating analysis cycles that previously required multiple rounds of manual research and stakeholder back-and-forth.
Platform and architecture reasoning
Used AI as a thinking partner on CMS architecture decisions, Drupal module evaluation, WebSphere integration strategy, and the executive narrative for the headless transition — faster and better-reasoned decisions than working alone.
The framing I use when asked about this: AI didn't replace the work. It raised the floor on how fast I could get to a well-reasoned first draft — for content, for strategy documents, for requirements, for executive narratives. The skill was knowing which problems were worth bringing to it, and being the human who reviewed everything that came out.
A CTO- and CMO-approved headless CMS architecture strategy with content hub design, COPE implementation plan, and acquisition migration logic — ready for team rollout when the engagement ended.
A regulatory intake workflow with SLAs that replaced informal "urgent" requests with a structured, accountable process — reducing friction, rework, and timeline disruption across teams.
Delivery templates and operating rhythms — intake, estimation, UAT, launch checklists, retrospectives — that improved quality and consistency across a high-volume release environment.
A documented AI workflow model: not just the outputs, but the framework for deciding when AI is appropriate, how to structure the human review layer, and how to preserve brand voice and compliance control throughout.
B2B ecommerce performance improvements through funnel separation, content alignment to sales goals, and UX improvements — plus the analytical model connecting web behavior to business outcomes.