What we do
We build and run the data layer that reporting sits on. Four practices, delivered through three stages: assess, build, run. Most clients start with an assessment and take one stage at a time.
PRACTICE ONE
Business intelligence and analytics engineering
- Semantic data models and star schemas
- Power BI and Microsoft Fabric development
- Tableau and Looker Studio where you already own them
- Report rationalization, fewer reports, each one trusted
- Pipelines on Azure SQL, Fabric, scheduled Python or dbt
PRACTICE TWO
Data governance and management
- Metric definition registers: one definition, one source, one owner
- Data quality frameworks and ongoing monitoring
- Source-of-record mapping
- Ownership and stewardship models, access reviews, retention policy
- Master data cleanup
- Regulatory-compliance readiness for analytics data
PRACTICE THREE
Advanced analytics, machine learning and AI
- Forecasting: revenue, demand, census, cash
- Segmentation, cohort and lifetime value analysis
- Propensity and risk scoring: churn, lapse, readmission, denial, conversion
- Experiment design with proper sample sizing and honest readouts
- AI readiness, the clean semantic model and documented definitions that make Copilot give you right answers
PRACTICE FOUR
Growth and marketing analytics
Market measurement and optimization.
- GA4 and tracking implementation
- Multi-touch attribution
- Acquisition cost and lifetime value payback modeling
- Channel and campaign return measurement
- Funnel, conversion and retention analysis
HOW THEY FIT TOGETHER
Four practices, one sequence
Governance finds the problem. Business intelligence builds the fix. Growth measures what it earns. Machine learning predicts what happens next. Governance keeps all of it correct.
You do not buy a practice. You buy a stage and the stage draws on whichever practices your situation needs. That is why the assessment comes first: it establishes which of these you actually require, rather than asking you to guess.
DELIVERY
How an engagement runs
01
Assess
A fixed-scope assessment of your data estate. Where the numbers break, which system owns which figure, what you can and cannot currently measure. Delivered as audit findings with the cost of each defect, plus a prioritized remediation roadmap. Read-only access; no changes to any production system.
$2,500
02
Build
The findings register from the assessment defines the scope, so the build is specified by evidence rather than estimate. Integrations and pipelines, a modeled data layer, master data cleanup, the metric definitions register, and documented lineage so every number can be defended. Priced by number of systems integrated.
from $6,000
03
Run
An ongoing engagement in which we operate your analytics and reporting layer. Dashboards built and maintained as the operation changes. The monthly and quarterly management pack produced on time and reconciled. Regulatory, payer, funder and investor submissions handled. Data quality monitoring.
from $2,500
FIT
Who this is for
Organizations of roughly five to 250 people, running at least two or three systems of record that do not reliably agree, with no analyst or BI developer on staff. If everything you run is in spreadsheets, there is nothing to reconcile yet. If you cannot obtain read-only access to your systems, or cannot sign a data processing agreement, we cannot deliver the work.
Read-only by default
Assessments run on read-only extracts. No changes to any production system.
Insured
Professional liability and cyber liability cover in force. Certificate on request.
Contracted properly
Mutual NDA, MSA, per-engagement SOW, DPA. Business associate agreement where HIPAA applies.
Fixed price, fixed date
You approve a scope and a number. Changes go through a written change order.
Find out what your data can and cannot tell you
Not ready to talk? Score your own data estate first. The 30-Point Data Health Checklist. Get the checklist