FrameRead

From close reading to editorial work

A source-tracked system for turning rigorous close reading of television and film into reusable editorial forms.

Relationship
Studio development
Stage
In development
Capabilities
  • Product strategy and management
  • Applied AI
  • Software and data engineering
  • Research and content systems

Availability

The method is being tested outside the studio. Transcript preparation and editorial production do not replace verification against the original work.

The problem

Close reading becomes difficult to reuse when transcript preparation, analytical claims, comparisons, essays, and scripts each develop their own unsupported version of the work.

The contribution

JodyInk created a canonical analytical record and source-tracked production workflow that can support several editorial outputs without losing the basis of the reading.

Decisions that shaped the work

  1. Label AI-inferred transcript attribution as provisional and preserve raw material when enhancement fails.

  2. Derive downstream trackers and editorial formats from a canonical record, with drift checks between representations.

What was implemented

  1. Chunked transcript preparation with provider fallback, explicit attribution metadata, and failure preservation.

  2. Canonical analysis records that support trackers, comparisons, essay inventories, scripts, and short-form planning.