A platform modernization and integration story for a $7.8M ARR business - vendor evaluation, integration architecture, telemetry, and governance.
Every disconnected tool created its own data silo, its own manual handoff, and its own risk of drift between systems. Leadership had no reliable, real-time view into throughput, adoption, or where quality was slipping, because the data needed to answer those questions lived in six different places at once.
The business needed to keep growing, but the operating model underneath it couldn't scale with more headcount alone - adding people would only staff around the fragmentation, not remove it. The real fix had to be structural: one platform, one data model, and integrations that kept every system in sync automatically.
I owned strategy, roadmap, and rollout for the platform end-to-end. I led vendor evaluation across SaaS platforms and APIs using a weighted scoring framework covering data model fit, integration depth, extensibility, access controls, and total cost of ownership, then drove the executive selection and directed migration off six legacy tools.
I consolidated 12+ core business workflows and shipped 10+ integrations (Dynamics GP, Adobe Sign, Azure AD, Power BI, SQL) via APIs and event-driven pipelines, creating a single source of truth for downstream reporting and finance. I then defined a four-layer KPI and telemetry framework (efficiency, adoption, quality, business throughput) and instrumented executive and operational dashboards in Power BI, Azure DevOps, and SQL to guide prioritization and release sequencing.
Three credible paths existed for consolidating six legacy tools. The right call depended on integration depth and long-term ownership cost, not just which option looked fastest on paper.
| Option | Data Model Fit | Integration Depth | TCO | Verdict |
|---|---|---|---|---|
| Keep Point Solutions, Integrate Ad Hoc | Poor, fragmented across silos | Low, manual bridges only | Low upfront, high long-term cost | Rejected, doesn't fix root fragmentation |
| Build a Custom Internal Platform | High, fully controlled | High, but built one integration at a time | High build & maintenance cost | Rejected, too slow, ties up engineering |
| End-to-End SaaS Platform, Open APIs | Strong, unified data model | High - 10+ systems integrated via APIs & events | Predictable, usage-based | Selected |
We selected the end-to-end SaaS platform: it gave the fastest path to a single source of truth, with enough extensibility via APIs and event pipelines to integrate 10+ existing systems without taking on custom-build risk.
Scored SaaS platforms and APIs on data model, integrations, extensibility, access controls, and TCO to drive the executive selection.
Directed migration off six legacy tools onto the unified platform.
Consolidated 12+ core business workflows into the new system.
Shipped 10+ integrations (Dynamics GP, Adobe Sign, Azure AD, Power BI, SQL) via APIs and event-driven pipelines.
Instrumented the four-layer KPI framework and kept high-risk approvals human-in-the-loop.
Rollout is part of product, not just engineering - a great platform launched badly is a failed platform.
Visibility is a feature. The four-layer KPI/telemetry framework changed how leadership prioritized, not just what they could see.
Adoption depends on workflow trust, not raw functionality - teams forgive missing features faster than broken trust.
The platform decision had real disagreement behind it - some stakeholders pushed for a custom build for maximum control. I brought the weighted scoring framework - data model fit, integration depth, TCO - into the room, and we converged on the SaaS platform once the tradeoffs were quantified instead of argued qualitatively.
I tracked cycle time and throughput weekly against the pre-migration baseline throughout rollout, not just at the end - catching a workflow regression early enough to fix before it hit go-live metrics.
Not every good idea ships at once. Deprioritizing advanced reporting to protect core transactional workflows, splitting large releases to cut regression risk, and keeping high-risk approvals human-in-the-loop were deliberate scope calls that protected trust and compliance over speed.
