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

NNW Clip Engine · Oct 2025 – Present

Turn a live signal into a reviewable clip while it still matters.

An event-driven operating surface for detecting broadcast moments, creating clips, preserving context, and handing finished media to publishing.

Ownership
Sole designer, engineer, and operator
Primary users
Media operators who need to catch fleeting live moments and move them toward distribution without repeating the same manual handoff for every clip.
Film 01 · 27 sec · silent

Four views from the live Clip Engine: channel control, the clip log, operator commands, and the social publisher.

The problem

The work starts before the interface.

A strong live moment can disappear before an editor sees it. Manual clipping, metadata entry, review, and platform publishing add delay exactly when relevance is highest.

6,000+followers reached by automated accountsReported Sep 2026
300,000+total likes across published contentReported Sep 2026
4publishing destinationsYouTube · TikTok · Instagram · X

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 Clipbot control room showing three monitored NNW channels and operator controls
Figure 01

Operate live capture by channel

The production control room keeps each monitored channel, bot state, and manual intervention point visible without exposing the detection implementation.

  • Per-channel state makes operational responsibility explicit.
  • Manual controls remain available beside automated monitoring.
NNW Clipbot activity log with public Twitch clip links, triggers, channels, and queue actions
Figure 02

Preserve a structured event record

The live activity view pairs each public Twitch clip with its trigger, channel, operator, timestamp, and next queue action.

  • Source links and trigger provenance travel with the asset.
  • Queue entry is an explicit handoff rather than a hidden side effect.
NNW Clipbot command reference showing production clip commands and usage guidance
Figure 03

Make intervention learnable

The command surface documents the operator path for creating and managing clips while a stream is live.

  • The workflow is discoverable inside the tool.
  • Command guidance reduces reliance on private runbooks.
NNW social publisher showing destination selection, clip input, and the studio queue
Figure 04

Hand the clip to publishing

The production publisher separates source selection, destination choice, generated copy, and final delivery state.

  • Editors can review the asset before a platform action is taken.
  • Destination-specific work shares one visible queue.

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. Live signals
  2. Event worker
  3. Clip and archive
  4. Review queue
  5. Publisher adapters

Scope

What the system owns.

  1. 01

    Combines chat velocity, audio events, spoken phrases, and operator commands as candidate signals.

  2. 02

    Creates Twitch clips, archives underlying media, and records structured metadata.

  3. 03

    Routes completed clips into Discord notification and social review queues.

  4. 04

    Generates destination-aware titles, captions, and hashtags before distribution.

Results

What changed.

  • Reduced the distance between a live moment and a reviewable publishing asset.
  • Powered automated content accounts to 6,000+ followers and 300,000+ likes.
  • Kept operator review available while automating capture, archive, and routing work.

Technology

Built with.

Python, TypeScript, Twitch API, Discord API, FFmpeg, PostgreSQL