Skip to case study
Aryan ChaubeyNNW / Media IntelligenceField record · 2025—26
← NNW platform

NNW Creator Intelligence Platform · Oct 2025 – Present

Find momentum before audience size tells the story.

A continuously refreshed research surface for searching the public Twitch ecosystem, spotting rising creators, and inspecting relationship context.

Ownership
Sole designer, engineer, and operator
Primary users
Creator scouts and operators who need to move from a broad public catalog to a defensible shortlist without losing the signals behind each result.
Film 01 · 32 sec · silent

Four views from the live Creator Intelligence product: Explorer, Rising, graph context, and the relationship canvas.

The problem

The work starts before the interface.

Creator discovery is fragmented across profiles, categories, and momentary rankings. At ecosystem scale, both the scoring work and the browser delivery model have to remain useful under real memory and execution limits.

1,000,000+creator records indexedReported Sep 2026
~799,617matches in the captured Explorer viewLive product · Sep 10, 2026
11.8Kcategories represented in the graphLive product · Sep 10, 2026

Production interface

What the live product makes visible.

Each frame below is a direct capture of the operating product. The annotations describe the workflow decision the view represents.

NNW Creator Explorer showing search, filters, and approximately 799,617 public matches
Figure 01

Search the creator universe

The production Explorer combines a direct search field, progressive filters, and public profile results in one place.

  • The captured result count reflects the active public-data filter.
  • Filters stay secondary until a scout needs to narrow the catalog.
NNW Creator Intelligence Rising view with live public Twitch creator rows and radar scores
Figure 02

Rank momentum, not just size

The Rising view puts audience change, consistency, status, and public profile details next to the scouting actions they inform.

  • Radar scores surface candidates before raw follower count dominates.
  • Shortlist, watch, and compare actions remain attached to each result.
NNW relationship graph controls showing graph mode, categories, and public creator counts
Figure 03

Choose the graph context

A scout can inspect the global category map or begin with a reference creator before requesting a neighborhood.

  • The graph discloses its current sample and universe counts.
  • Reference-led exploration avoids loading unrelated edges.
NNW Creator Intelligence relationship canvas with creator and category nodes
Figure 04

Inspect relationships on demand

The graph canvas renders only the selected context, keeping a large public network navigable inside the browser.

  • Category and creator nodes share one explorable canvas.
  • Edges are computed only for the requested neighborhood.

System boundary

A readable flow, not a deployment map.

This is the level of architecture useful to a product reviewer: responsibilities and handoffs, without schemas, endpoints, thresholds, or topology.

  1. Public creator signals
  2. Ingestion and enrichment
  3. Scoring layer
  4. Discovery index
  5. Scouting workspace

Scope

What the system owns.

  1. 01

    Continuously ingests and enriches public creator signals.

  2. 02

    Computes recommendation scores, relationship graphs, and rising-creator signals.

  3. 03

    Supports shortlists, watchlists, comparison, and repeated scouting over time.

  4. 04

    Delivers a large creator universe with packed typed arrays, chunked responses, lazy-loaded facets, and on-demand graph edges.

Results

What changed.

  • Brought search, momentum signals, creator profiles, and graph context into one research workflow.
  • Kept large-scale exploration within serverless execution and browser-memory limits.
  • Made public creator signals legible as ranked candidates rather than a raw directory.

Technology

Built with.

PostgreSQL, Python, Next.js, TypeScript, React