The Problem
A Series D fintech had accumulated the classic BI stack problem: Sigma, Power BI, Tableau, and Looker—all running simultaneously. Different teams had different tools. Nothing was integrated.
The real pain? Dashboard requests sat in a queue. Someone needed a new view of customer data? Submit a ticket. Wait. Follow up. Wait more. Weeks later, maybe you'd get your dashboard.
The analytics team had become a bottleneck. Not because they were slow—because demand for data insights far outstripped their capacity to build dashboards.
The Approach
We didn't try to unify the BI tools. We replaced the entire workflow with what we call "Dashboards in Meetings."
The old process:
Before
- Submit a dashboard request
- Ticket enters the queue
- Analyst picks it up (eventually)
- Back-and-forth on requirements
- Build in BI tool
- Review, revise, repeat
- Finally delivered (weeks later)
After
- 15-minute meeting
- AI spins up a plan
- Double-check together
- Build with AI assistance
- Validate and ship
- Same day delivery
The New Workflow
Schedule a 15-Minute Meeting
The person who needs the dashboard books 15 minutes with a BI engineer.
Discuss Requirements Live
No more async back-and-forth. They talk through exactly what they need, together.
AI Spins Up a Plan
Claude generates a dashboard specification based on the conversation.
Double-Check Together
Both people review the AI's plan. Catch misunderstandings before building.
Requester Can Leave
Once the spec is approved, the requester's job is done. No more waiting.
Build with AI Assistance
The BI engineer builds the dashboard with AI, validating as they go.
Deliver with Documentation
Dashboard is sent back with a "map"—documentation that explains how to use it.
Iterate If Needed
Changes needed? Just ask. The cycle time is so short that iteration is painless.
The Result
The key insight: this isn't about replacing the BI team with AI. It's about human + AI collaboration that makes the human massively more effective.
The BI engineer is still involved at every step. They validate the AI's plan. They catch edge cases the AI misses. They ensure data quality. But the grunt work—the SQL queries, the chart configuration, the documentation—that's accelerated by 10x.
"BI is now an afterthought of the company. It's so easy. It's freed up so many people's work—they can do other things now."
What used to take weeks now happens in hours. More importantly, the analytics team stopped being a bottleneck. They went from constantly behind to actually having bandwidth for strategic work.
The transformation:
- Queue time: Weeks → Same day
- Requirement gathering: Async tickets → 15-minute meetings
- Build time: Days → Hours
- Iteration speed: Painful → Trivial
- BI team capacity: Bottleneck → Strategic asset
The Bigger Lesson
This isn't a BI story. It's a workflow story.
Every company has bottlenecks like this. Teams that are underwater because demand exceeds capacity. Queues that exist not because the work is hard, but because there's too much of it.
AI doesn't replace those teams. It multiplies them. The BI engineer who could build one dashboard a day can now build five. The same expertise, distributed further.
Is Your Analytics Team a Bottleneck?
Let's talk about what "Dashboards in Meetings" could look like for your organization.
Book a Call