Building Zasper: Inception to Growth

Zasper started out as an open-source project focused on expanding and improving the capabilities of Project Jupyter. This talk reports on how the project has evolved in terms of adoption.

Many FOSS talks focus on how to get your first 10,000 users. This talk is about what happens when you build something highly specialized and discover your user graph isn't wide—it’s deep. Zasper (a high-performance, Go-backed Jupyter IDE) secured a FOSS United grant and initially saw high broad adoption post-launch. However, as the initial curiosity tapered off and those general numbers went down, a unique reality emerged: low broad adoption, but extreme usage intensity from a core cohort of power users. I'll share the architectural decisions that kept those power users hooked, the realities of balancing independent contracting with open-source stewardship, and why retention is the ultimate validator for a niche developer tool.

00:00 - 00:05 | The Itch & The Architecture

1. The core thesis: Why standard Jupyter setups struggle under heavy logs, massive datasets, and complex rendering.
2. Choosing Go for an interactive notebook IDE backend—leveraging its low-overhead concurrency and runtime efficiency to fix the performance gap.

00:05 - 00:10 | Fueling the Project

1. Securing the FOSS United grant: Moving from a personal side project to a community-backed initiative.
2. Launching: Taking a step back to focus on what’s actually needed

00:10 - 00:18 | The Power-User Paradox (Core Section)

1. Deconstructing the post-launch metrics: Facing the reality of high to low raw download counts.
2. Discovering the "Deep Graph": Spotting the intense, multi-hour daily usage patterns of a dedicated cohort.
3. Deep Dive: The precise engineering choices (state synchronization, low-latency communication via goroutines) that made the tool indispensable to these power users.

00:18 - 00:22 | Pragmatic Stewardship

1. The balancing act: Managing the demands of open-source stewardship
2. Shifting community strategy: Why prioritizing deep feature stability for 14,000 power users matters more than building a funnel for 50,000 casual clickers.

00:22 - 00:25 | Conclusion & Live Demo

1. A lightning-fast live demo showcasing Zasper running smoothly under heavy stress loads.
2. Final takeaway: Why retention, not virality, is the true indicator of product-market fit in developer tooling.