Cineshots.AI shot list
0→1 AI product Film

Cineshots.AI

An AI tool that turns film scripts into shot lists. Used on productions for SHOWTIME, Prime Video, and Tubi. Users report up to 90% less pre-production time.

Role

Lead Product Designer

Timeline

1 year

Platform

Web app

Worked with

CEO and dev team

TL;DR

Problem

Directors of Photography build shot lists by hand. It takes dozens of hours per script and invites errors.

My Role

Lead Product Designer (0→1), partnering with the CEO and dev team on workflows and the first release.

Constraints

Early-stage AI (Jan 2023), film-set realities, and a hard requirement to keep humans in the loop.

Impact

Used on SHOWTIME, Prime Video, and Tubi productions. Users report up to 90% less pre-production time.

Context and my role

CineGach Studios brought me on in January 2023, when AI film tools were just emerging. I led product design from first concept through launch and six months after, with the CEO and dev team.

I had scored films, mixed sound on set, and directed my own shorts. I knew the workflow, which let the DP interviews go deeper.

The problem and the evidence

A shot list is the shooting schedule and the on-set bible. DPs build them by hand, dozens of hours per script, with plenty of room for error.

Plenty of pre-production tools existed. None were built around the cinematographer's process.

We spent three months testing concepts with working DPs before any production code. {{TODO: the key finding from DP testing}}

The bet

AI drafts the shot list and the DP keeps control of every shot.

Suggestions are labeled as suggestions, editable, and traceable to the script page they came from. The bet was that this would earn enough trust for real productions.

Success measure: {{METRIC: success measure for the bet, e.g. share of AI shot suggestions kept or lightly edited}}.

Options and trade-offs

Automation vs. control

We kept manual overrides everywhere so cinematographers trusted the AI. This cost extra UX complexity but preserved creative ownership.

Model accuracy vs. latency

We prioritized faster script parsing with progressive loading, accepting lower initial detail in exchange for speed.

Feature depth vs. onboarding

We trimmed advanced features from the first-time flow to keep activation under 10 minutes.

What shipped

Day planning with recommended equipment

Results

Up to90%
Less pre-production time, user-reported
<10 min
First-time activation
{{METRIC: productions or active users}}
Adoption

After launch we spent six months refining the product on data from working filmmakers and their real productions.

What I'd do differently

{{TODO: my reflection}}