FivePaths

Television Academy Interviews

Screenshot of Television Academy Interviews Homepage on desktop
interviews.televisionacademy.com
Client
The Emmys

The Television Academy Foundation’s The Interviews brings television history to life through in-depth conversations with people in front of and behind the camera. Its public website connects those recordings with the shows, people, professions and subjects discussed in them. Behind that experience is MASH, the internal catalogue the archive team uses to manage records and send content to the public site.

FivePaths rebuilt both Drupal sites together, preserving established public URLs and the working practices librarians relied on. The project combined migration, site configuration, custom development and testing with the client. Subsequent work has extended MASH with tools for reviewing transcripts alongside video, while our ongoing support covers maintenance, bug fixes and improvements as the archive’s needs change.

Two sites, one publishing workflow

The separation between MASH and the public Interviews website was an established part of the client’s workflow. The archive team liked that arrangement and wanted to retain it through the rebuild. Our task was to carry it forward on the rebuilt platform, with room to improve individual tools and interactions.

That meant understanding the catalogue as a set of connected records. An interview can have several recorded parts. Clips identify passages within those parts, and catalogue relationships connect them with shows, people, topics, professions and genres. On the public website, those relationships give visitors several routes into a long conversation. Someone interested in a particular show can find the passages about it alongside the full interview.

Diahann Carroll interview with the Shows tab selected and clips grouped under programme names

An interview can be explored through its shows, topics, professions, genres and people. This view groups Diahann Carroll’s clips by show. Current public interface, captured November 13, 2026.

Mobile Diahann Carroll interview page with video, wrapped catalogue tabs and the first show grouping

On a narrow screen, the interview’s catalogue tabs wrap above the show-linked clips. Current public interface, captured November 13, 2026 at a 390 by 844 CSS-pixel viewport.

The connection between the sites carries catalogue records and their relationships from MASH into Interviews. It needs to recognize the corresponding public records when editors export changes, so that an update reaches the right place in the archive. Maintaining that connection was part of rebuilding the two sites together.

Within FivePaths, Tyler Shuster handled development, Jason Salter configured the site, Eric Leland managed the client relationship, and Beth Brockman helped with testing and QA. The librarians’ knowledge of their catalogue and everyday tools helped the team check the rebuilt workflow against real work.

Moving the archive together

The most demanding part of the migration was moving both sites in tandem. Public visitors already knew the Interviews website, and editors had tools and routines they used throughout the catalogue. The team had freedom to change the implementation while preserving those points of continuity.

Keeping existing public URLs was a core requirement. We checked Drupal’s Pathauto patterns against titles as part of preserving the archive’s addresses. The implementation also included redirect handling for older links that carried an additional /interviews path prefix. Link behavior received attention alongside the content migration.

Testing followed the route content actually takes through the system. We asked the archive team to work with a small, varied set of interviews: older records, newer records and detailed examples using the available fields. They could make typical edits in MASH, export the records and check the corresponding content on the public site. Repeating that process with different records exposed issues across the boundary between the two systems.

One small omission showed why the librarians’ participation mattered. On the old publishing screen, a Title search let them find records to send to the public site even when those records had not recently changed. They used it regularly and flagged its absence from the rebuilt Export screen. Jason restored the search during testing, and the client confirmed the adjustment. A familiar tool for selecting records belonged in the new workflow.

The public Interviews site relaunched on August 19, 2025. Follow-through included correcting a record-identifier mismatch that affected interview content and working with Pantheon to restore MASH search indexing. Moving the sites together also meant checking their behavior together as they settled into use.

Making catalogue details easier to inspect

Catalogue quality depends on details spread across many records. A misspelled name can be difficult to spot when it appears among otherwise valid field values. Librarians needed a way to compare what had actually been entered across the catalogue.

We developed a field-audit tool that lists the distinct values stored in a field, shows how often each appears and links to the records using it. Staff can compare spellings and follow a questionable value back to its source. The tool provides a view across records to support inspection and correction.

Field History addressed a related question: when did a particular field change? Drupal already held the underlying data in its revisions. We developed the Field History module to make that existing information easier to see at the field level, presenting changes to a selected field over time.

The two tools provide complementary views of the catalogue. Field audit brings together values from multiple records. Field History follows a field through the saved revisions of one record. Both make stored information accessible in a form that helps staff investigate a specific detail.

Reviewing transcripts alongside the video

Later work built out MASH’s transcription capabilities, with a separate phase of development and client testing in 2026. The work connected generated text, speaker information and video timing within the archive’s existing records.

AssemblyAI supplied speaker identification and word-level timestamps. It also accepted names and keywords from MASH to guide transcription. We connected that input with information the catalogue already held: interviewer and interviewee names, related people and shows, and keywords supplied for an interview or an individual part. That gave the transcription service context specific to the recording, including names it might otherwise struggle to recognize.

The interface needed as much attention as the service integration. Long transcripts have to remain practical for a person reviewing the recording. Performance and natural interaction were central concerns, including how reviewers move between a passage of text and the moment it represents in the video.

In MASH’s rendered transcript, clicking a word takes the player to that point in the recording. The interface pairs speaker-labelled passages and timecodes with a movable video player, and highlights words as playback progresses. Speaker-reassignment and search-and-replace controls support further work on the text. These interactions keep the transcript and recording connected as someone checks a passage.

Diamond Farnsworth’s video beside speaker-labelled transcript passages and timecodes in MASH

MASH pairs a movable video player with speaker-labelled, timecoded dialogue. Clicking a word seeks to its moment in the recording. Current authenticated interface, captured November 13, 2026.

The keyword work continued through testing and rollout in 2026. It used vocabulary from the archive to guide generated text while keeping the transcript available for human inspection. AWS storage supports the transcription data, and Pantheon hosts the Drupal sites.

Continuing to care for both sites

Since the relaunch, FivePaths has continued routine maintenance, bug fixes and improvements requested by the client. The platform has also moved to Drupal 11, with testing and follow-up work across both sites.

The continuing relationship includes maintaining the connection between MASH and Interviews as well as extending individual tools. New requests still need to fit the archive team’s publishing practices and the public site’s content relationships. That was central to the rebuild and continues to guide the work as the archive develops.

Built Using

Drupal

Venerable open-source CMS for managing large sites efficiently

Pantheon

Seasoned host with a variety of developer-friendly features

YouTube

World's largest video hosting provider that can be embedded and integrated

AWS

Highly flexible cloud services provider

Bootstrap

Ubiquitous and extensible styling framework that accelerates site design

Docker

Container management for developing reproducible tech stacks

Gitlab

Powerful self-hosted code version control software with workflow integration

PHPStorm

Indispensable development environment for all our web projects