Flowstate AI company logo - AI Video Intelligence Platform

How a Motorsports Team Cut Editorial Time by 95% with Flowstate

Case Study

8min Read

A leading motorsports organization used Flowstate to make a 60-terabyte archive searchable, automate tagging, track sponsor exposure, and generate race highlights in minutes instead of hours.

A leading motorsports organization spent years sitting on thousands of hours of race footage with no efficient way to use it. Finding a single clip could take an editor half an hour. Producing one highlight video took three to four hours. Sponsor reporting was slow, incomplete, and hard to defend.

Over a multi-week pilot, the team tested Flowstate’s AI video platform across their entire content operation. By the end, work that used to take hours was taking minutes.

What changed

  • The 60-terabyte archive became fully searchable in plain language

  • Every clip was automatically tagged by driver, sponsor, venue, and event type

  • Highlight videos were generated directly from live broadcast feeds

  • Sponsor exposure was tracked at the frame level, with timestamps

About the Customer

The organization is one of North America’s top motorsports competitors. They field multiple drivers across two major racing series and work with more than 30 commercial sponsors.

Their marketing team publishes 3 to 5 short videos every day across Instagram, TikTok, YouTube Shorts, X, and Facebook, reaching an audience of over 4 million followers.

Every race weekend produces a large volume of footage:

  • In-car cameras on every driver

  • Live broadcast feeds

  • Drone footage

  • Audio from crew chiefs and spotters

  • Behind-the-scenes video from the garage and shop

Over the years, this has added up to a library consisting of more than 60 terabytes of footage.

The Challenge: A Content Operation Running on Human Memory

“We had hundreds of clips every week and a 60-terabyte library going back years. The footage was incredible. Our problem was that the only way to find anything was for a person to remember it and scrub for it.”

— Senior Director of Marketing

After every race, footage from cameras and field crews flowed into folders with names like “11 pit stop” or “11 push to grid.” Editors then watched and tagged each clip manually. With hundreds of clips coming in every week, this created four compounding problems.

1. Tagging took half of every Monday

Every clip had to be tagged by hand: driver, sponsor, location, event type. This took half of Monday plus 3 to 4 additional hours later in the week. The more races, the more time it consumed.

2. Finding old footage was slow and unreliable

The archive ran on filename search across 60 terabytes of storage. Finding one specific clip, say a pit stop from a particular driver at a particular race for a particular sponsor, could take an editor over 30 minutes. Footage from earlier seasons was rarely used at all.

3. Sponsor reporting was incomplete

The team had obligations to more than 30 commercial partners, all of whom needed proof that their logo appeared on screen. Their existing tool tracked logo appearances but missed brief ones, could not provide exact timestamps, and required a manual setup process for each new logo that needed to be onboarded.

4. Most footage never got used

Broadcast rules limited how much race footage the team could post on social media. With editing time as the main constraint, they typically produced one highlight video per race win. Out of 100+ hours of available footage each weekend, roughly 5 hours made it to publication.

The Solution: What Flowstate Did for the Team

Flowstate was tested across four parts of the content operation: finding footage, tagging clips, reporting sponsor exposure, and creating highlight videos.

1. Search the archive in plain language

Before Flowstate, finding a specific clip meant remembering roughly where it was and scrubbing through files manually. With Flowstate, editors typed requests in plain English:

  • “Early-morning shots at the shop”

  • “Semi-truck on the way to the track”

  • “Victory-lane celebration with family on stage”

Flowstate returned results with timestamps and thumbnail previews, linking directly to the moment in the source video. It surfaced clips editors remembered but could no longer find, and several they had never seen at all, including brief sponsor logo appearances that would have been missed at normal playback speed.

2. Tag every clip automatically

Flowstate analyzed footage frame by frame and applied tags based on a structure the team defined:

  • Who: drivers, cars, racing series

  • Where: venue, content type such as pit stop, burnout, garage, and B-roll

  • What: visible sponsors, clip titles, descriptions, and hashtag suggestions

Accuracy started in the mid-80s and reached the high 90s after the team provided a small set of reference images for their eight drivers and top sponsor logos. The Monday-morning tagging session was removed from the calendar entirely after utilizing Flowstate’s AI video intelligence.

3. Track sponsor exposure at the frame level

The team ran a direct comparison between Flowstate and their existing sponsor tracking tool, using one major sponsor across the same content window:

Metric

Existing Tool

Flowstate

Posts with sponsor visible

49

52

Total accumulated views

1.0M

2.2M

Exact timestamp per appearance

Not supported

Supported

Cost to add a new sponsor logo

$150 per logo

$0

Time to onboard a new sponsor

6 to 8 weeks

Same day

Flowstate also picked up spoken mentions of the sponsor in broadcast audio, something the existing tool did not track. Sponsor reports could now include timestamps, structured data, and archived clips that a partner’s team could verify themselves.

4. Produce highlight videos in minutes, not hours

Previously, turning a four-hour race broadcast into a highlight video took 3 to 4 hours of editorial work. With Flowstate, an editor provided the broadcast and a short description of the desired story. Flowstate identified the key moments: victory-lane celebrations, in-car reactions, family moments, and crew radio highlights. The output was a 60 to 90 second video, already formatted for vertical platforms like TikTok and Instagram.

The first highlight took five minutes of editor time from start to finish. Per race win, output went from one highlight to four: an expert pick, a fan favorite, a family cut, and a sponsor-focused version.



The Results

“We could go from a four-hour broadcast to a publishable highlight in the time it used to take us to find the right clip. And it gave us sponsor reporting we could actually defend to a partner with timecodes.”

— Director of Digital Media

Across the pilot, the team measured improvements in four areas.

Time Saved

Task

Before

After

Weekly clip tagging

3 to 4 hours

Effectively zero

Producing a highlight reel

3 to 4 hours

5 minutes

Sponsor analytics review

2 hours manual

30 minutes automated

More content per race

  • Highlight videos per race win went from 1 to 4+

  • Sponsor reporting went from post-level presence to exact frame-by-frame timestamps

Better sponsor detection than the existing tool

  • 6% more posts with sponsor presence identified

  • 120% more accumulated views attributed to sponsor exposure

  • Frame-level timestamps and spoken mentions added, neither of which the previous tool supported

Lower operational costs

  • Sponsor onboarding went from 6 to 8 weeks down to the same day

  • Per-logo onboarding cost dropped from $150 to $0

  • The 60-terabyte archive became fully searchable for the first time

Why This Matters Beyond Motorsports

The challenge this team faced is not unique to racing. Any organization that produces large volumes of video, whether from match days, live events, or regular programming, runs into the same wall. The footage exists. The value is there. But without the right tools, most of it never gets used.

Manual workflows force a choice between doing things well and doing them quickly. Sponsor reporting built on screenshots and slow setup cycles cannot keep up with a real partner portfolio. And an archive that can only be searched by someone who remembers what is in it is not really an asset at all.

Flowstate removes that bottleneck. It turns raw footage into searchable, timestamped data and lets small editorial teams produce at a volume that would otherwise require significantly more staff or mean leaving content unused.

How Flowstate Can Help Your Organization

“It started out as one thing, just tagging, and now it’s so much bigger.”

— Director of Marketing, Motorsports Organization

If your team manages large volumes of video from live events, races, or regular programming, Flowstate can help you:

  • Find any clip in your archive using plain-language search

  • Tag footage automatically by person, location, sponsor, and event type

  • Track sponsor logo appearances at the frame level, with timestamps

  • Generate highlight videos directly from broadcast or field footage

  • Produce platform-ready social content without additional editing tools

  • Connect to your existing cloud storage and editorial workflows

Book a Demo.

About the Author

Sahil Shah

Founder & CEO, Flowstate

Sahil Shah is the Founder and CEO of Flowstate. Prior to founding the company, he spent nearly a decade at Waymo and Apple building large-scale video AI and computer vision systems, and has over 15 years of experience bringing frontier video technologies from research into production environments.

About the Author

Sahil Shah

Founder & CEO, Flowstate

Sahil Shah is the Founder and CEO of Flowstate. Prior to founding the company, he spent nearly a decade at Waymo and Apple building large-scale video AI and computer vision systems, and has over 15 years of experience bringing frontier video technologies from research into production environments.

Testimonials

Read Similar Case Studies

Case Study

How Attention Economics Quantified Editorial Bias Using AI

Case Study

iFit Case Study: Scaling video metadata tagging and content discovery with Flowstate AI.

How iFit scaled video metadata tagging with Flowstate

Case Study

South Park Commons Case Study: Automating social media clip generation from long-form video.

How South Park Commons Scaled Social Content from Long-Form Video

Experience FlowState in action

Explore Enterprise-Grade Video Intelligence Built for Scale and Security.

Experience FlowState in action

Explore Enterprise-Grade Video Intelligence Built for Scale and Security.

Experience FlowState in action

Explore Enterprise-Grade Video Intelligence Built for Scale and Security.