Price: 0
Number of applications: 1
10.12.25 (inclusive)
Non-cash
MVP
ICT tasks
Media sphere
Neurotechnology and artificial Intelligence
Software/ IS
Difficulties in building an implementation roadmap (PoC → Pre-Prod → Prod), misunderstandings in the deployment process, caching weights and version control, as well as difficulties in describing the interaction between modules: assistant (LLM/RAG), scene shaping unit and generative models.
- Detailed deployment plan for Wan 2.2 and Flux 1 models on servers. - Technical specification for the implementation of the inference pipeline. - An architectural diagram of the entire ML chain (from texts to videos). - Recommendations on infrastructure, resources, optimization and scaling.
Gorchkhanov Timur Tagirovich
Purpose and description of task (project)
The project combines two areas: Deployment and optimization of generative models (such as Wan 2.2 and Flux 1) on our servers. Designing the architecture of the entire chain, in which the AI assistant analyzes text documents (briefs, descriptions of scenes, characters) and transmits the results to visual models.
Whatsapp
