Production companies spend thousands of costly man-hours ingesting, sorting, annotating and reducing raw footage for further post-production stages. Especially within documentary films where the average shooting ratio between raw footage and final film is 80 to 1. For feature films it takes an average of 16 weeks to edit 1 hour of final film. Existing industry software falls short in providing an effective one stop solution that fits the needs of Post producers and editors.
Our solution is to offer an A.I. powered search engine that specifically caters to our users’ needs and can be optimized and customized for a specific project . This tool will save up 60% of time spent in the log, loading and pre editing phase.
One of our future clients (one of the world's most successful feature documentary film production companies), explains how our product can help her business “(The xxxxx (brand name) tool) ... would make a huge difference to our daily editing process and it would allow directors and editors to experiment more and let them focus on artistic expressions.“
The company will eventually be widely available for not only production companies, but also to individuals, featuring a range of new innovative features like SCENE SUGGESTION and SENTIMENT ANALYSIS through an ecosystem that plugs in to industry standard software like Avid Media Composer and Premiere Pro.
Business model and Go-To-Market
The business model for our service will roll out in two stages. Focusing in the beginning on bespoke B2B solutions with internal human resources, and then gradually converging to include a B2C model that features a fully automated editing assistant.
The first stage targets larger film -and TV Broadcasters and larger cultural institutions with a heavy data load. Secondly, we will target specific independent film and TV productions companies. Here we will offer a full-service package where we hire a human assistant editor (validator) to quality check our machine's job and help it train on the specific footage.
The price for this service will be less than what it would cost the production to hire their own assistant editor, and we expect to engage customers turning an average monthly revenue per project of 2500 €.
After this initial stage our MVP will have been tried and tested from real world projects and customer needs and will then be ready to roll out in stage two under a flexible billing plan for both B2C and B2B. Here we expect that the costs of inhouse human validators have decreased substantially and ultimately, they will be more or less removed from the internal costs.
Price structure in stage 2:
“Pay as you go”- Free subscription, after 10 gb of material 0.1€ per minute.
“Creative Individual” - 25 € per month, 0.05 € per minute/gb
“Studio” - 350 € per month per project, 0.04 € per minute/gb
“Enterprise” Individually negotiated. (500-50.000 € a month) Focused on providing extensive customer and technical support, as well as adding missing features and custom ML models.
The solution is first and foremost a tool for film and TV makers, directors, editors, editor assistants and producers. Their interest in our product has been validated through 25+ in depth interviews and through a wider quantitative survey distributed through international filmmaker communities. 82% of filmmakers confirmed interest.
Secondarily we have identified a need for our services with broadcasters, film studios and film archives. Here the volume of audio-visual production (and storage of it) is much higher, and content formats are less varied, which makes it easier for our tool to help on a larger scale.
We currently have a close to home client that has requested our technology in a high-profile film production. A contract with the production company has been signed and the film is set to go into post-production in 2021.
We are also in close negotiations with the following hand-picked customers in relation to our strategic approach: Blu, Final Cut 4real, United, Museum2Go.
The Museum2go case came to us through an EU consortium (Applying for a EU innovation grant) that wants us to develop our technology to automate video marketing films for European museums. Within a year we would develop deployable technology for an online video marketing platform and launch it together with a business model that would generate revenue in form of licencing, maintenance cost and per data fee’s processed by clients.
Perry Nightingale, Senior Vice President of WPP, the world's largest marketing company that produces approximately 25% of all commercial messaging globally, has this to say about Kaspar AI; 'If you are able to speed up or improve video production using AI you will be very successful. Lockdown is going to change filmmaking and AI can help.'
According to a rapport from DR (The Danish public national broadcaster) and the DFI (Danish Film Institute) over 2000 hours of documentary and TV productions were released in Denmark in 2018. The average wage of an editor according to the Danish Filmworker Union is €2150 a week (+facilities). An average editing time for producing 1 hour of final TV production is 7 weeks (16 weeks for features), accumulating an average editing cost of €13.790 per final film production.
This amounts to a current market spend on editing of 42 mill euro - in Denmark alone.
For the first stage of our go to market strategy (B2B) our marketing will revolve around film and TV tradeshow and festival presence, free online AI editing workshops for key customers and targeted online user acquisition.
Additionally, our strategy is to deploy our AI in a high-profile international feature film that will reach the attention of film festivals, international press and ultimately film and TV professionals. We currently have a client that has requested our technology in such a production. A contract with the production company has been signed and the film is set to go into post-production in 2021.
This would ideally coincide with our roll out for B2C where our marketing would revolve around more aggressive online campaigns, ambassadors promotion and community based awareness.
We are competing with an array of AI powered tools for the cinema industry, like Drylabs, Valossa, Silverstack, iconic.io and axle.ai.
Furthermore, we are competing with existing editing software like Adobe Premiere and Avid Media Composer, that are constantly developing and releasing new tools for their existing user base.
What makes us different is the fact that we through extensive user research have identified a range of key features that are missing in current search solutions; like CINEMATIC IMAGE FEATURES, ACTIONS, CHARACTERS and SPECIFIC OBJECT/LOCATION RECOGNITION. Additionally we have found that there is a need for intuitive TIME and CONTEXT aware search options including NATURAL LANGUAGE integration.
The team has financed the development since the beginning of 2019 with the help of a range of grants from The Danish Film Institute and CEFIMA in Norway (approx. DKK 500.000). Since April 2020 the company has been financed by the Innofounder program, of which 3 founders have been admitted. (approx. 1.700.000 DKK).
1.500.000 DKK is now being raised through Innobooster to complete the company's MVP.
CEO - M.A. Film producer (36 years) Previous app startup founder and independent cultural entrepreneur, responsible for raising 15+ million DKK in project support.
Product Owner - (32 years) MA. Architect and filmmaker, 5+ professional experience in the film industry and extensive cinematic knowledge.
Lead Designer - (32 Years) Programmer, multidisciplinary artist and graphic designer. 5 + years experience designing AI solutions at a major German corporate industry leader.
CTO - (25 years) Machine Learning engineer, graduated from Medialogy from AAU at the top of his class specialized in AI and automatic media creation.
Strategic Business Developer (34 years) Cand. Tech. Master in Technology (Operation and Innovation), with 6 years experience as business developer in a larger media publishing companies.
Spørgsmål til panelet
How do we actually start generating recurring revenue from enterprise customers at the early stage, and how do we grow our customer base?
o Later, how do we pivot from B2B to B2C in a sustainable way, retaining customers and their business with us?
Using humans to assist our AI is “cheating” and false marketing, but is it not necessary to secure quality results for customers?
o Furthermore, later when humans are taken out of the equation, how do we sustain “human” feel to our product and services?
We believe that we need to take the lead and show the industry the potential of AI in audio visual storytelling, therefore we are focusing a large part of our energy in realizing an international feature film. Is this a good idea or not? Why?
o Furthermore, how do we balance R&D in relation to business growth?
How do we draw a line between commercial possibilities versus ethical questions on whether an AI should have authoritative control of media creation?
o And how do we manage the suspicion and skepticism with AI and the fear that it will take jobs?