Project Snowbird

Context

Project Snowbird was a future-vision design initiative focused on a foundational product question: as Snowflake expanded beyond its original workflows and users, what should Snowsight become? The team saw that Snowsight was no longer scaling cleanly. Workflows felt disconnected, capabilities were hard to discover, and the product experience did not yet reflect the breadth of the platform.

The opportunity was to rethink Snowsight as more than a collection of tools. Snowbird explored how it could become the gateway to Snowflake's broader ecosystem and help drive discovery, adoption, and a more cohesive experience across the platform. The project was not about polishing a few screens. It was about defining a clearer direction for how Snowsight should scale, what role it should play in Snowflake's ecosystem, and how design could make that future tangible.

Approach

Snowbird combined internal research, synthesis, competitive analysis, user framing, and concept exploration. The team interviewed internal stakeholders, translated findings into opportunity areas and how-might-we questions, and explored multiple future models instead of locking too early into one answer.

The work moved through five phases: framing the problem (Snowsight was not scaling to the breadth of workflows and users Snowflake now needed to support), exploring the design space broadly before committing to one direction, grounding the work in stakeholder research and synthesis, turning strategy into concepts and prototypes, and packaging the direction for leadership, design org review, and planning conversations.

One of the strongest structural ideas to emerge was a more intentional information architecture: top navigation for global structure, a contextual left panel for projects and assets, and a right panel for details on demand. The team also explored how AI could improve the experience, but treated it as an amplifier behind the scenes rather than a standalone feature.

Research

Snowbird set out to answer three questions:

• What and who is Snowsight for?
• What are the untapped opportunities in the product experience?
• How might Snowsight unlock more value for Snowflake and its customers?

Research simplified Snowsight's audience into four user types:

• Core Customer — people for whom Snowflake's core offerings are central to day-to-day work
• Extensibility Customer / Partner — people unlocking new business value from Snowflake's newer offerings
• Non-User Customer — people whose companies value Snowflake but who don't yet engage with it directly
• Newcomer — people still learning what Snowflake is and whether it fits their needs

That framing pulled the work beyond the assumption that Snowsight only needed to serve deep technical experts—and made room for onboarding, education, feature awareness, and broader pathways into the platform.

Internal stakeholder interviews surfaced recurring themes: feature discovery was a top pain point; users often struggled to find the right data; the product architecture exposed visible seams instead of hiding complexity; and the team needed to reduce unnecessary complexity rather than add more of it. AI came up repeatedly, but stakeholders wanted it as a behind-the-scenes amplifier—not a standalone feature users had to learn separately.

Directions

The team evaluated materially different operating models for Snowsight—not just screen iterations.

Snowsight Plus was the closest to an evolved version of today: all-in-one, context-driven, and visually cleaner, with complexity hidden until needed. It felt familiar and could reduce visible clutter, but risked repeating a jack-of-all-trades problem and hiding capabilities power users needed to discover.

Product Suite organized the experience by persona or job-to-be-done, with mode-switching between workflows like analyze, develop, or administer. It could tailor the experience to user mindset, but real work often spans multiple modes—mode-switching could add friction without strong anticipation or automation.

Omakase was the most ambitious model: a highly anticipatory experience that adapts to patterns of use, surfaces what needs attention, and shows actions taken on the user's behalf. It pushed toward the most differentiated future, but depended on a high degree of trust and strong escape hatches if the system guessed wrong.

The team also explored mobile concepts, universal features like scratchpad and suggestions, graph views of object relationships, and storyline-based end-to-end flows across ingest, code, debug, manage, monitor, and consume.

There was also a practical tradeoff between future vision and near-term product constraints. The team explicitly split attention between Snowbird as a future-facing effort and immediate left-nav quality-of-life improvements as steps toward that vision. Even in the final storytelling, the IA direction was framed as an opportunity space rather than a hard prescription, so teams would not feel prematurely locked into one structure.

Role

I was project lead on Snowbird—not sole owner of every deliverable, but responsible for shaping direction, narrative, and team alignment. My work sat at the intersection of concept design, systems thinking, and project shaping.

I explored several early future-facing concepts in low fidelity—including Snowsight Plus, Product Suite, mobile-oriented ideas, and anticipatory patterns like suggestions, scratchpad, and graph views of object relationships. I also explored the Omakase concept, which pushed further into anticipation, trust, and machine-supported momentum.

Beyond broad concepting, I contributed specific experience areas that helped make the vision more concrete: a monitor and debugging hub with data quality representation, a refined editor experience, and a more consumer-feeling business mode concept. I created working input for the monitoring experience and drove parts of the competitive analysis that argued for top navigation, a contextual left panel, a details-on-demand right panel, and scalable accordion menus.

I participated in stakeholder interviews, helped surface recurring themes, and brought a clear point of view around discovery, organization, reducing visible product seams, and using AI as an amplifier rather than the main event. I helped articulate the end-to-end journey across ingest, code, debug, manage, monitor, and consume—and worked to strengthen the project story, align the team around a clearer structure, and contribute to the leadership-ready presentation package.

Impact

Snowbird's impact was strategic rather than feature-level. The project produced a clearer narrative for Snowsight's future, grounded in research and design, and gave leadership and partner teams something concrete to react to. More broadly, it helped establish a north-star direction for Snowsight: a more scalable, discoverable, and cohesive product experience that could better reflect Snowflake's expanding platform and user base.

There are no direct revenue, adoption, or satisfaction metrics attributable to Snowbird—and I would not claim those externally. What is measurable and defensible is explicit research and alignment output: completed interviews, synthesis that informed conceptual iteration, strategic artifacts meant to shape decisions, and evidence that stakeholders felt heard.

Most of Snowbird remained visionary work: repositioning Snowsight as the gateway to Snowflake offerings, designing for a broader set of users, a more intentional IA model, and AI as an embedded amplifier. Some ideas appear to have influenced nearer-term navigation and IA thinking as incremental quality-of-life improvements—baby steps toward the vision rather than the full vision itself.

The strongest signal of influence beyond the project team is that Snowbird was later referenced as a design-org north star in company all-hands and influenced how Snowsight vision was discussed in leadership and planning contexts—north-star influence more than direct shipped business impact.