System Integration at @Imagemaker
Overview
A fast-growing services company was facing a common scaling challenge: operational and financial information was scattered across several unconnected tools (CRM, ATS, onboarding, HRIS, ERP). Because there was no single source of truth, leadership struggled with delayed reporting, duplicated manual work, and inconsistent KPIs. As Integration Lead, I designed and implemented a full data consolidation layer that unified all core business processes into a dimensional model. This empowered executives with reliable, real-time insights on revenue, delivery capacity, staffing, and financial forecasting
Business Impact & Key Achievements
- Achieved a 20% workload reduction in Change Management area by aligning system integrations with strategic goals and automating redundant processes
- Enabled real-time tracking of billable headcount and service assignments, which was critical for financial forecasting and operational planning.
- Created a single source of truth for core indicators: projected revenue, active assignments, margin by market, and staffing utilization.
- Unified data pipelines served as the foundation for other strategic initiatives, including MOCA.
Technical Approach
The solution combined API integrations, scalable data modeling, orchestration automation, and executive dashboarding.
API Integration Layer (Python)
I implemented a series of Python-based connectors to extract structured data from CRM, ATS, HRIS, ERP, and onboarding systems. Data was normalized and stored in a staging schema for transformation.
Dimensional Modeling (Kimball Principles)
I designed a star schema around the company’s core business processes, including: Dimensions Client / Account Employee / Talent Opportunity / Engagement Product / Service Offering Country & Date Dimensions Facts Projected revenue Billable headcount Staffing assignments Opportunity lifecycle Monthly revenue projections The model supports slicing by market, account, seniority, profile type, and date.
Automated Orchestration
To replace manual data maintenance, I implemented: Scheduled serverless jobs running extraction and transformation scripts Workflow orchestration in Azure Logic Apps ensuring dependencies execute in the correct order Automatic updates to downstream analytics models This guaranteed fresh data for decision-making every day
Executive Dashboards (Power BI / DAX)
Using the dimensional layer, I built dynamic dashboards that provide: Capacity vs demand Margin insights by account and service Headcount utilization Forecast vs actual revenue Assignment lifecycle analytics
Challenges & Learnings
Mapping internal vs. external IDs across systems required collaboration with business owners. Handling incomplete records (e.g., missing maker_id_client_partner) led to building fallback logic. Writing SQL views that were both performant and understandable by analysts was key for adoption
Strategic Tradeoff: Why We Chose Logic Apps Before a Custom Integration Layer
I chose Azure Logic Apps over a custom integration engine because it allowed us to deliver value immediately. Logic Apps gave us a reliable orchestration layer with retries, scheduling, monitoring, and dependency management out of the box — letting us focus our engineering time on data modeling, Python extractors, and the KPI layer instead of building infrastructure. This decision reduced engineering effort, lowered risk, and accelerated time-to-value during a period when leadership urgently needed unified financial and operational visibility. By deploying integrations in days instead of months, we enabled real-time headcount, margin, and revenue forecasting much earlier in the project lifecycle. The fast iteration cycle also let us validate transformations, understand system load, and refine business rules before considering a long-term custom solution. The tradeoff was clear: optimize for speed and impact now, and evolve to custom orchestration later with real usage data — a pragmatic technical decision that delivered measurable business outcomes.