PLATFORM // ARCHITECTURE

Development Platform for Film & TV: Stories backed by verifiable comparables.

A learning system for filmmakers, producers, studios, broadcasters and streaming providers. Stories and creative concepts are analysed, matched against open data sources and consolidated into structured data sheets. Every session enriches the knowledge base. Data delivers the findings, people deliver the judgement.

01 // MODULES

Four modules, one continuous system.

Each module resolves a concrete bottleneck in daily development work. Together they form a closed working cycle that becomes more precise with every session.

Asset Library Icon

MODULE A

Asset Library

The curated knowledge store for your organisation. Stories, talent, creatives and external IP observations in a structured, searchable wiki dashboard.

Living Sheets Icon

MODULE B

Living Sheets

Intelligent data sheets that grow with every session. Autonomous initialisation from verified sources, version history and chat feedback loop in one pass.

Intelligence Chat Icon

MODULE C

Intelligence Chat

Natural-language queries across the entire data corpus. Every figure carries its source, collection date and confidence interval. No claim without a reference.

Data Engine Icon

MODULE D

Data Engine

An extensible collection of open data sources and connections to social APIs, platforms, communities and industry registers. Provider-neutral and always interchangeable.

MODULE A // KNOWLEDGE STORE

Asset Library: the central wiki dashboard

No more scattered Word documents and untraceable email attachments. The Asset Library acts as the curated knowledge store for all your stories, talent, contacts and external IP observations. Story and concept files can be created, classified and shared with the entire development team.

Every entry follows a type-specific schema. Fields for rights status, casting options, format parameters and market positioning are not free text but structured fields with defined values.

  • Multi-dimensional filters: format, micro-genre, rights status, territory
  • Full-text and semantic search across 80-150 curated micro-genre tags
  • Automatic duplicate detection for author and title names
  • Audit trail for every record: who changed what and when
EXPLORER // ASSETS (842 ENTRIES) INDEXING: LIVE
> filter: region:DACH rights:optionable format:series 14 RESULTS
Alpine Noir Ensemble Drama Village Satire Office Thriller +71
TITLE / ENTITY CATEGORY TAGS STATUS
Der Schatten uber dem Inntal Literature / IP #Alpine-Noir OPTION OPEN
Nachtwache Spree (Series) Original Project #Crime #Berlin IN PITCH
Elena Vogel (Showrunner) Creative #Historic #South AVAIL Q1/26
FILTER ACTIVE: DACH TERRITORY SHOWING: 1-3 OF 842
LIVING SHEET // ASSET-LIT-402 V4.1 - AUTONOMOUSLY GROWN

IP TITLE

Das Schweigen der Tundra (Thriller)

VERIFIED
Rights Holder Hanser Verlag (Agency)
Sales Copies 74,000 (Hardcover)
Audience Sentiment 4.7 / 5.0 (Reading platform)
Micro-Genre Tags #Polar #Isolation
Comparables Expeditionen 2019 / ARD
Funding Compatibility FFF Bayern: HIGH
FEEDBACK FROM CHAT SESSION #3911 3 HOURS AGO

New fact extracted: "Option for adaptation expired 31.03.2026 without renewal. Agency has signalled willingness to discuss (agency email, verified)."

SOURCE: Internal chat session • CONFIDENCE: 91% • DATE: 14.09.2026
MODULE B // INTELLIGENT FILES

Living Asset Sheets: data sheets that keep pace

No data sheet ever goes stale. The moment you create a story or talent entry, an autonomous agent pulls the base data from verified sources and formats it consistently. The flywheel principle takes hold here in practice: every session enriches the sheets, the system improves with each run.

Every conversation your team has in the chat about an asset flows back as a verified fact into the data sheet. The system grows with every development session.

Autonomous initialisation on creation

AI pre-fills base data from verified sources

Continuous re-sync

Every change versioned with a timestamp

Chat facts feed back in

Conversation results written in with confirmation step

MODULE C // LANGUAGE INTERFACE

Intelligence Chat: with complete provenance

Ask complex editorial questions across your entire data corpus in natural language. The agent decides in the background: do the internally held data suffice, or does a targeted market research query need to be triggered?

Every figure carries its attribution: source, collection date and confidence interval. No claim without a reference. AI models are provider-neutral and interchangeable, including locally hosted models.

SOURCE + DATE + CONFIDENCE
INTELLIGENCE CHAT // SESSION #4092 PROVENANCE: STRICT

QUERY // DEVELOPMENT:

"Which showrunners in the Europe have demonstrable experience with historical material and are still available in the second half of 2026?"
STRATEGY: HYBRID (INTERNAL + OPEN INDUSTRY SOURCE) DURATION: 1.1s

4 high-match profiles identified. Full dossier from page 2.

Johannes K. (Berlin) CONFIDENCE: 96%

Most recent: Head Writer on a 6-part broadcaster event series (1920s). Agency confirms availability for next project from August 2026.

SOURCE: OPEN INDUSTRY SOURCE DATE: 28.05.2026 NO RETENTION
ALL FACTS ARE WRITTEN BACK INTO THE RESPECTIVE ASSET SHEET UPON CONFIRMATION.
> Ret

DATA SOVEREIGNTY // EUROPE-FIRST

EU hosting. Local models. Your data stays yours.

CineCockpit runs on EU infrastructure and is designed for GDPR compliance from the ground up. The AI layer is provider-neutral: local models can be used, with no dependency on any single vendor. Screenplays, treatments and unpublished concepts never leave your instance.

EU HOSTING

Located in Germany / EU

All instances run in EU data centres. No data transfer to third countries without your approval.

INTERCHANGEABLE MODELS

Provider-neutral

The AI model is configurable and replaceable. Local inference supported. No lock-in.

GDPR

Zero retention for research queries

External research queries are not stored permanently. Data processing agreement available on request.

02 // TAXONOMY

5 Asset Types. Fully modelled for the industry.

Each type carries a data schema aligned with the real conventions of the European film and television industry.

TYPE // 01 PERFORMANCE

Talent

Actors, comedians, creators and influencers with relevant audience transfer. Profiled by audience segment, dialect competency and reach data from open platform sources.

EXAMPLES // DACH: Ensemble casts, emerging award winners, dialect profiles, audience reach correlations.
TYPE // 02 CREW & DIRECTION

Creatives

Directors, screenwriters, showrunners, producers and key cinematographic figures, captured with filmography, genre profile and availability.

EXAMPLES // DACH: Writers room leads, award recipients, genre specialists, co-production-experienced practitioners.
TYPE // 03 BOOK MARKET

Literature / IP

Novels, non-fiction, graphic novels, stage plays and verified life rights. With sales data from DACH book market sources and open work and metadata registers. Stories are matched against this database.

EXAMPLES // DACH: Bestsellers, indie discoveries, true crime monographs, historical memoirs.
TYPE // 04 NEW FORMATS

Other IP

Games, storytelling podcasts, web and short-video formats, audio dramas, music catalogues: formats absent from conventional development databases.

EXAMPLES // DACH: Narrative podcasts, German indie games, investigative research series.
TYPE // 05 DEVELOPMENT PIPELINE

Projects (Original Stories)

Your own stories in development: from the two-page idea sketch through treatment and screenplay to the finished pitch deck. Screenplay and material analysis runs directly within the system; results feed into the project file. Connected to talent, market and funding profiles.

FUNCTION: Linked to talent and market profiles, automatic package review, funding history.

03 // OPERATING MODES

Three modes for every step in the development cycle.

Whether you are scanning the market for fresh impulses, reviewing a submitted package or preparing a detailed negotiation.

01

Discovery

RADAR FOR OPPORTUNITIES

A continuous radar across the German-language book and rights market. Discover newly published titles, award-winning stage plays and yet-unoptioned IPs before they are offered to major broadcasters.

-- Trend early detection without hype distortion
-- Weekly curated scouting reports
-- Automatic availability notifications
02

Validation

ASSESSMENT OF A PACKAGE

A book or screenplay has been submitted to you. Enter the package. CineCockpit cross-references tone, casting suggestions and market environment against prior DACH productions.

-- Check of competing stories in on-demand libraries
-- Profile matching of requested creatives
-- Risk factors in the funding market
03

Deep Mining

6 DATA BLOCKS PER ASSET

Comprehensive in-depth analysis of a specific asset. Generates a structured 6-part matrix: rights position, track record, distribution chain, audience segments, funding compatibility, comparables.

-- Granular fact verification per data point
-- Exportable dossier (PDF / Markdown)
-- Direct handover into the internal chat

WORKFLOW // PROCESS OVERVIEW

Schematic of the agentic research process: query, source selection, resolution, findings

04 // CORE WORKFLOWS

Four workflows development teams use every day.

From a quick comparison query to a full treatment import.

WF // 01

Comparison query in the chat

You type an editorial question: "Which similar stories aired on a major public broadcaster in the last 3 years, and what ratings did they achieve?" The agent searches your library first, then supplements from verified DACH media library data and delivers the result with complete provenance.

INPUT: Free text OUTPUT: Table + source citation DURATION: < 2 seconds
WF // 02

Deep Asset Mining (6 data blocks)

For a specific asset, a complete 6-block analysis is generated:

01Work Basis
02Commercial DE
03Audience / Sentiment
04Creator Brand Analysis
05Adaptation Comparables
06Competition / Pipeline
INPUT: Asset ID or name OUTPUT: 6-block dossier, exportable
WF // 03

Market query: demand / supply gap and white space

Targeted market analysis for a genre or topic. The system checks which stories are currently present in Europe on-demand libraries (supply), where audience interest lies (demand from sentiment and search data) and flags the resulting white space as a concrete opportunity. Stories are cross-referenced against open sources and social platform signals.

INPUT: Genre / micro-genre tag OUTPUT: Gap matrix + source citation
WF // 04

Material input: upload a logline or treatment

Upload a logline, treatment or outline. CineCockpit runs an automatic screenplay and material analysis: micro-genre classification, identified comparables, potential audience segments and initial research pointers are extracted and staged for your review.

CONFIRMATION STEP BEFORE SAVING:
All extracted fields are presented to you for review. No record is saved without your approval.
INPUT: Logline / treatment (PDF, TXT) OUTPUT: Structured story profile CONFIRMATION: Required

WORKFLOW // END-TO-END STREAMLINE

The complete AI streamline: from concept and screenplay through analysis to marketing.

Six stages. One continuous process. Data delivers the findings, people deliver the judgement.

INPUT

Idea, Story and Screenplay

Free text, treatment, screenplay or concept as the starting point.

ANALYSIS

Story Profile and Screenplay Analysis

Extraction of genre, tone, characters and format parameters.

MATCHING

Databases, Social Channels and Communities

Structured matching against open sources and audience signals.

PROPRIETARY AI

Mistral (European AI)

EU infrastructure, interchangeable and self-hostable. No third-country dependency.

FINDINGS

Market Findings and Provenance

Comparables, audience match and cited sources with confidence.

OUTPUT

Marketing Approach

Recommended approach for digital channels, derived from market findings.

05 // GO-TO-MARKET

From story to audience: a data-informed marketing approach.

Alongside the story analysis, CineCockpit produces a data-informed marketing approach: relevant audience segments, suitable digital channels from social media and community platforms, timing indicators and comparable campaign patterns from the European market. Everything is presented as a finding and a proposal. The decision stays with the team.

Data-backed forecasts, no guarantees, no invented reach figures. The marketing approach is a structured starting point for the creative and strategic team.

MARKETING CHANNELS // OVERVIEW ILLUSTRATIVE
Schematic of digital marketing channels: social media, community platforms, timing patterns for film and TV
ILLUSTRATION // Channels and timing patterns based on market findings. Data-backed forecast, no guaranteed ROI.
01

Audience segments

Drawing on comparables and audience signals, CineCockpit identifies concrete audience segments for the story: age, region and thematic affinities from aggregated, non-personal resonance data.

02

Digital channels and communities

The system identifies the digital channels and thematic communities where comparable stories generate the strongest resonance. Social media platforms, forums, fan communities and streaming discussions are evaluated.

03

Timing indicators and campaign patterns

Comparable campaigns from the European market show which phases and windows have been used for similar titles. Patterns from publicly accessible campaign data, no proprietary media planning data.

04

Marketing findings in the data sheet

The marketing approach for a story is written into the relevant Living Sheet. All recommendations carry a source reference and confidence level. Nothing is saved without your confirmation.

DATA-BACKED BOX-OFFICE GUIDANCE. NO AUTOMATIC GREENLIGHT PROMISE. DATA DELIVERS THE FINDINGS. PEOPLE DELIVER THE JUDGEMENT.
MARKETING FINDINGS // STORY: NACHTWACHE SPREE (SERIES) ILLUSTRATIVE

AUDIENCE SEGMENT

25-45, DACH, crime-oriented, urban

Derived from audience resonance of comparables Dark (2017) / Kleo (2022)

SUGGESTED CHANNELS

Social video, crime communities, newsletter formats

Resonance of comparable titles strongest on video platforms and thematic forums

TIMING PATTERN

Autumn window (Oct-Nov), 6-8 weeks lead time

Pattern from European campaign data for comparable series launches (publicly accessible)

ALL FIGURES ARE PROPOSALS BASED ON AGGREGATED DATA. DATA-BACKED FORECAST. NO GUARANTEED ROI. DECISION RESTS WITH THE TEAM.

06 // DATA ENGINE

Data sources, signals and agentic workflow.

CineCockpit connects to an extensible collection of open data sources and interfaces linking social APIs, platforms, communities and databases. All connections are provider-neutral and interchangeable. The agent selects precisely from the available source pool for each question, resolves entities via open knowledge graphs and returns results with source citation, date and confidence. No black-box score, no automatic greenlight.

METADATA

Film/TV Metadata

Cast lists, filmographies and international identifiers from publicly licensed work and metadata registers.

API // PUBLIC
ENTITY HUB

Knowledge Graph

Entity resolution via open knowledge graphs and linked data registers. Foundation for clean attribution without duplicates.

LINKED DATA // CC0
DEMAND SIGNAL

Search Interest

Aggregated demand and reach signals from public platforms and community data sources. DACH-weighted, time-series capable. No personal data tracking.

SIGNAL // AGGREGATED
RESONANCE

Audience Signals

Audience resonance for soundtracks, trailer reach and streaming activation for book adaptations from public sources.

SIGNAL // PUBLIC
STREAMING DEMAND

Streaming Indicators

Streaming demand indicators for DACH and the EU from public platform data. Comparables research for film production and licensing.

DATA // EU FOCUS
BOOK MARKET DACH

DACH Book Market

Sentiment data, sales figures and bestseller lists for literary IP in the Europe from open book trade databases.

DACH PRIMARY
CREW & TALENT

Industry Database

Specialised industry networks for film professionals in the Europe. Filmographies, availability and genre profiles from open industry sources.

DACH PRIMARY
ON-DEMAND LIBRARIES

Public Broadcaster Libraries

Public production data, broadcasts and on-demand library presence of public broadcasters in DE, AT and CH. Competitive and market analysis for DACH.

PUBLICLY ACCESSIBLE

The agentic workflow at a glance.

STEP 01
Natural-language question

Question in plain language: story, genre, comparison, market situation.

STEP 02
Agent selects sources

Automatic selection of the relevant interfaces and data sources based on the question type.

STEP 03
Corpus + entity resolution

Full corpus, no sampling. Entity resolution via open knowledge graphs.

STEP 04
Findings with provenance

Result with source, collection date, confidence and cost transparency per run.

GUARDRAIL: NO AUTOMATIC GREENLIGHT // DATA-BACKED BOX-OFFICE GUIDANCE DATA DELIVERS THE FINDINGS. PEOPLE DELIVER THE JUDGEMENT.
Schematic of interconnected data sources and signals in the CineCockpit engine

ILLUSTRATION // Network of open data sources, signals and entity registers. The engine selects precisely from the available source pool for each query.

07 // CHAT FLOW

How the Intelligence Chat works.

You ask a question in plain language. In the background an agent selects the appropriate tools and sources, matches the data and returns an answer in which every statement is backed by a source, a collection date and a confidence level. No claim without a reference.

INTELLIGENCE CHAT PROVENANCE: STRICT
Which comparable titles to an alpine thriller have aired in Europe since 2021?
AGENT // TOOL SELECTION
1 Search internal library
2 Fetch on-demand library data
3 Check audience resonance from open sources
Data matching in progress... 847 entries checked. 9 results. Entities resolved.
FINDINGS // 3 COMPARABLE TITLES
Bergschatten (6-part series, public broadcaster, 2022)

Micro-genre: Alpine Noir / Ensemble Drama. Strong audience resonance in AT and southern DE.

SOURCE: Public library data DATE: 03.06.2026 CONFIDENCE: 94%
ALL FACTS CAN BE WRITTEN BACK INTO THE STORY DATA SHEET ON REQUEST.
01

User question in plain language

Describe your research goal the way you would explain it to a colleague. No code, no filter syntax.

02

Agent selects tools

The system decides automatically: internal library, external open sources or both. Every step is traceable.

03

Data matching and entity resolution

Titles, people and properties are unambiguously resolved via open knowledge graphs. No guesswork, no duplicate problem.

04

Answer with provenance

Every statement carries its source, collection date and confidence level. Results can be written directly into the Living Sheet of the relevant asset.

GUARDRAIL: NO CLAIM WITHOUT A REFERENCE. NO AUTOMATIC GREENLIGHT.

08 // TECHNOLOGY

Technology and European AI.

CineCockpit is built on an AI layer that was developed and is operated entirely in Europe. As our language model we work with Mistral, an AI built in France and one of the leading European language models. Models are interchangeable and can run locally or self-hosted, with no cloud dependency on providers outside the EU.

MISTRAL // EUROPE-FIRST

Mistral as the default model

As our language model we work with Mistral, an AI developed in Europe. Mistral AI is headquartered in Paris and provides models that can be operated on EU infrastructure. The models are well suited to multilingual content across European markets.

  • Developed and maintained within the EU
  • Operated on EU servers with no data transfer to third countries
  • Zero retention for research queries

Interchangeable model layer

The AI layer is designed to be fully interchangeable. Mistral is the default model. On request, other models including locally hosted open-source variants can be used without any architectural changes. No vendor lock-in.

Mistral (default) Self-hosted Local / on-premise EU cloud

GDPR-COMPLIANT

Data protection is not a compromise

External research queries are not stored permanently on the model side. Unpublished stories, treatments and screenplays never leave your instance. Data processing agreement available on request.

EU HOSTING

Infrastructure from Germany and Europe

All instances run in data centres in Germany or the EU. No unsupervised data outflow to third countries. Location is selectable and can be fixed contractually.

DATA SOVEREIGNTY

Your data belongs to you

CineCockpit runs as a dedicated instance per organisation. No multi-tenant data mixing, no use of your data for model training outside your instance.

TRANSPARENCY

No black-box score

Every system response carries its source, date and confidence. The model does not invent facts. It processes what you bring in and what comes from verified sources.

AI LAYER: INTERCHANGEABLE // DEFAULT: MISTRAL (EU) // HOSTING: EU DATA CENTRE EUROPE-FIRST. ZERO-RETENTION. GDPR-COMPLIANT.

09 // RESONANCE DATA

Data matching across social media and communities.

Stories and topics are not only checked against film and book market databases, but also against signals from social channels and thematic communities. The result is a complete picture: what is present on platforms and in audience communities, what is being discussed and how people respond.

DATA BASIS

Open databases and registers

Stories are matched against work and metadata registers, book trade databases, funding databases and industry registers. All based on publicly accessible or licensed sources.

RESONANCE SIGNALS

Audience resonance from social channels

Aggregated signals from publicly accessible social channels and thematic communities show how strongly a topic or IP is in conversation. No personal raw data, no individual tracking. GDPR-compliant.

MATERIAL ANALYSIS

Bring in existing material

Ideas, treatments, outlines and screenplays are analysed and matched against the data basis. CineCockpit identifies comparable works, audience signals and market gaps based on the concrete material you submit.

FLYWHEEL PRINCIPLE

Every session enriches the data sheets.

CineCockpit learns from every research session. New facts developed in the chat flow back into the Living Sheets of the relevant asset. With every session the knowledge base becomes more precise and more complete. The system grows with your organisation.

  • Facts flow back after a confirmation step
  • Every change versioned with a timestamp
  • Nothing is written in without your approval
  • Library becomes more precise with each session
NO PERSONAL RAW DATA // NO INDIVIDUAL TRACKING // GDPR-COMPLIANT AGGREGATED SIGNALS. STRUCTURED SOURCES.

10 // SCREENPLAY AND CONCEPT REVIEW

Placing screenplays and concepts in context.

CineCockpit does not only evaluate market data and external IPs. It also contextualises your own creative concepts and screenplays. Ideas, treatments, outlines and finished scripts are analysed, matched against comparable works and positioned within the market. The output is a set of findings, not a verdict: data delivers the findings, people make the call.

01

Extract story profile

The system reads your material (logline, treatment, outline, screenplay) and automatically extracts a structured story profile: genre, micro-genre tags, setting, casting type, format parameters. All fields are presented for your review before anything is saved.

02

Find comparable works

CineCockpit matches the profile against the full library and verified European production data. Comparables are ranked by tone, genre proximity and format type. Every match carries its source and confidence level.

03

Audience match

Based on audience signals from the identified comparable works, CineCockpit shows which audience segments could be relevant for the submitted material. The matching is based on aggregated, non-personal resonance data.

04

Market findings and development stage

The system identifies where your material sits in the development process (idea, treatment, outline, screenplay) and provides pointers on market gaps, funding compatibility and the competitive pipeline across Europe.

STORY ANALYSIS // MATERIAL IMPORT CONFIRMATION STEP REQUIRED

DEVELOPMENT STAGE

Treatment (22 pages)

Detected: dialogue scenes, character arc, 3-act structure

COMPARABLES (TOP 2)

Alpenfieber (2023) / Dunkelziffer (2021)

Genre proximity: Alpine Noir // Confidence: 88%

AUDIENCE

35-55, Europe, thriller-oriented

Signal from audience resonance of comparable works

NO AUTOMATIC GREENLIGHT. DATA-BACKED FORECAST. DATA DELIVERS THE FINDINGS. PEOPLE MAKE THE CALL.

NEXT STEP

See the platform with your own material.

In a 30-minute live session we demonstrate CineCockpit using one of your real research scenarios or an anonymised IP.

Request a demo