A billion rows of claims data.
Processed in under two hours.
A Life Sciences healthtech startup replaced Alteryx with Dataworkz as the data platform behind its customer-facing insights. Months of pipeline turnaround compressed into hours. $75K per quarter returned to the business.
What this firm runs on Dataworkz.
- Customer
- Life Sciences healthtech startup
- Industry
- Life Sciences, insights management for HCPs & payers
- What Dataworkz replaced
- Alteryx
- What it powers
- Customer-facing insight delivery to healthcare providers, payers, and life sciences brands
- Scale
- ~200 employees · billion-row workflows · 18 Tableau dashboards delivered to customers
- Deployment
- AWS Kubernetes · Dedicated SaaS · horizontal scaling
- Stack
- Databricks Delta Lake→Dataworkz→Snowflake→Tableau
- Status
- Live
Because of Alteryx issues, we had to stagger our data projects out, it would take months to provide the analysis to our customers. Now with Dataworkz, we can deliver insights to our customers in near-real time, as soon as we get the data feeds from our sources.
A working data platform with three failure modes the BI team could not engineer around.
Months to deliver analysis.
The team had to stagger customer projects because the platform could not run two analyses concurrently at the volumes their HCP partners were sending. Sales and CS were apologizing for delivery dates on the product the customers were paying for.
One failure meant restarting from zero.
When a transformation step failed mid-run, the only option was to restart from the beginning. Hours of compute thrown away on every error. No intermediate state to resume from.
Licensing + EC2 + core servers + IT ops.
The Alteryx footprint was a measurable quarterly line item. Adding HCP volume added cost faster than it added throughput. The bill went up; the team's velocity didn't.
The team had a working data platform. They needed a different one.
A 280-step transformation workflow. No pipeline code. Restartable from any step.
The team rebuilt the entire prescription-and-claims transformation workflow on Dataworkz. Same source data. Same downstream BI. A different platform underneath.
Source
Prescription + medical claims data refreshed on partner cadence.
Lake
Source of truth lives on the customer's existing data lake. No movement out of their environment.
Transform
Multi-format ingestion · structured transformation · business-rule application · reference lookups · final dataset assembly. Each step observable, restartable, auditable.
Warehouse
Dataworkz writes to the customer's existing Snowflake. The BI team consumes those tables.
BI
Insight ships to the company's customers: providers, payers, life sciences brands.
Stay on Alteryx. Build it in-house. Or move to a platform that handles billion-row workflows out of the box.
Cost the team could measure, going up.
Continue staggering customer projects. Continue absorbing the licensing + EC2 + core-server + IT ops bill. Continue restarting pipelines from scratch on every failure.
280 steps live. 1B rows in under 2 hours.
Visual transformation workflow. Runs against the customer's existing Databricks + Snowflake stack. Scales horizontally on AWS Kubernetes the team already operated. Ships every workflow with end-to-end traceability and restart-from-step.
Six-to-twelve-month engineering bet.
Multi-step, observable, restartable transformations on billion-row data is not a quarter-long build. A platform team the company didn't have, against a delivery deadline that was already slipping.
The team picked Dataworkz because the platform shipped against the actual workflow: billion-row claims data, 280 steps, on the customer's existing data infrastructure, without requiring the BI team to become a platform engineering team to keep it running.
Months to hours. Quarterly TCO returned. Customer onboarding doubled.
| What they measured | Before | After |
|---|---|---|
| Time to process 1B-row prescription & claims dataset | Months | Under 2 hours |
| New data source onboarding | 6 weeks | 2 days |
| Quarterly TCO (licensing + EC2 + core servers + IT ops) | Baseline | $75K saved / quarter |
| Data refresh turnaround time | Baseline | 50% reduction |
| Customer onboarding throughput | 1× | 2× |
| Pipeline failure recovery | Restart from scratch | Restart from any step |
Outcomes from production deployment at this customer.
The data work behind your product can ship at the speed your customers expect.
If you're at a peer firm: life sciences insights, payer analytics, HCP data networks, clinical outcomes platforms, three things matter from this deployment.
The platform replaces a named tool.
Alteryx out, Dataworkz in. The transformation work that drove the decision was not a greenfield project, it was a working pipeline rebuilt on a new substrate. Brownfield migration is the realistic path for a BI team that already has a delivery deadline.
It runs on the data infrastructure you already have.
Databricks, Snowflake, AWS Kubernetes. No data movement, no new compute model to learn, no lock-in. The pieces the BI team already operated stayed in place.
Throughput shows up in the work your customers see.
Months-to-hours on the billion-row workload is the time your sales and customer-success teams stop apologizing for. The product gets faster on the outside because the platform got faster on the inside.
The data work behind your product is too important to wait six months to engineer in-house.
Book a working session with the Life Sciences team. We'll walk through the workflows in this case study against the data infrastructure your firm already runs on.