Contact person: Offline AI Animated Video Platform

Phone:Show

Email:Show

Location: Hamburg, Germany

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"Job Post: Build Full Offline AI Media Generation Platform – Hasaballa AI

- Only serious teams with proven AI experience should apply.

Platform: Hasaballa AI – Description
Hasaballa AI will run fully offline — even if it’s 200 GB in size.

Goal
Produce unlimited videos at high speed / top quality for Arab creators — no subscriptions, APIs, or internet. Quality must match (at minimum) these examples:
[login to view URL]
[login to view URL]
[login to view URL]
[login to view URL]
[login to view URL]
[login to view URL]

Workflow
1. Smart Director – Auto Mode
- User writes full script in Hasaballa GPT chat.
- Click “Smart Director – Auto.”
- System auto-generates everything (images, voices, animation, lip-sync, audio, final video).

2. Smart Director – Manual Mode
- Script written in chat.
- Click “Manual.”
- Script broken into steps: user decides generate/skip per stage (image, voice, animate, background, lip-sync).
- Each element added to final timeline.

3. Standalone Mode
- Each tool (Image, Voice, Animation, Lip-Sync, Audio) used individually.
- Final video assembly manual in this case.

*Future-Proof Requirement – Hasaballa Engine
Platform must support future integration with Hasaballa Engine (untrained core AI model).
It will learn from tools & user workflow, aiming to replace 90–95% of tools within a year, later split into sub-models (Text, Image, Animator, Voice) for online APIs.
All pipelines must be modular for seamless future integration.

---

*Hasaballa AI – 18 Core Components
1. System Requirements – Windows 11 (RTX 5090, 64GB RAM, 4TB SSD). Modular, future-ready.
2. Offline Execution – 100% offline: no APIs, no subscriptions.
3. Arabic Interface – RTL, single-page with all tools.
4. Chat Window – Central hub to write scripts, analyze scenes, insert reference images/voices, control Auto/Manual flow.
5. Image Generation – Ultra-realistic (16:9, 9:16, 1:1), cinematic lighting, identity lock, text-to-image, image-to-image, Arabic text support.
6. Image Animation – Full-body + face, lip sync, duration/camera control, cinematic FX (test SkyReels V2).
7. Voice Cloning & Generation – Studio-grade, emotions, pauses/breathing, 21 Arabic dialects + multilingual.
8. Background Audio – Built-in FX (emotions, war, nature, crowd, etc.) + optional Pixabay search (needs internet toggle).
9. Smart Director – Auto & Manual:
- 7-min video = ≤45 min (≈85 images, voices, animation, lip sync, sound).
- 15-min video = ≤90 min (≈170 images, full production).
10. Timeline & Final Video Editor (1–15 mins)
- Export: HD, FHD, 4K (16:9, 9:16, 1:1).
- CapCut-style: trim/split, transitions, filters, overlays, zoom/pan.
- Add external media.
- Subtitles/captions: manual/AI, multi-language, sync, style, translation.
- Drag & drop timeline with placeholders for skipped steps.
- Download, preview, layered timeline.
11. Smart Edit Layer – Edit using Arabic prompts with timecodes (replace visuals, sounds, etc.).
12. Integration – Google AdSense & YouTube (needs internet toggle).
13. Interface – 100% RTL, single-page, large buttons, tooltips.
14. Auto-Update Button (requires internet).
15. Auto-Save – Every 30s if changes occurred.
16. Project Folder System – Local storage for all assets per project.
17. Backup & Restore – Full backup of models, settings, projects.
18. Source Code Ownership – 100% exclusive to client, no reuse/resale.

---

*Preferred Team
- AI/ML Engineers
- Computer Vision/Animation Specialists
- Backend Developers (Python, FFmpeg, CUDA)
- Frontend/Fullstack (Arabic RTL UI)
- QA/System Optimizers
- Video Processing Experts

*Documentation
Full detailed document (tasks, sub-tasks, requirements) provided to the selected team.

*Budget & Milestones
Total budget: up to $7,500, divided equally across 6 milestones:
1. Task 5 – Image Generation
2. Tasks 7 & 8 – Voice Generation/Cloning/Enhancement + Sound Library
3. Task 6 – Image Animation & Lip-Sync (test SkyReels V2)
4. Tasks 4 & 9 – Chat Window + Smart Director
5. Tasks 10 & 11 – Timeline Editor + Smart Edit Layer
6. Tasks 12–18 – Final Integration + Full Source Code

*Bonus: $500–$1,000 for successful delivery at required quality.

---

General Conditions
1. Scope Integration – All milestones interconnected. No standalone tools.
2. Reference Quality – Must match or exceed reference videos. Any lower output rejected, developer must revise at no cost.
3. Client Approval – Each milestone requires client confirmation.
4. Unlimited Revisions – Within scope until desired quality met.
5. 100% Offline – Entire platform runs locally. Online only for Pixabay search & YouTube/AdSense.
6. Technology – Must use PyTorch, FFmpeg (optional: OpenCV, [login to view URL], CUDA). No third-party APIs, no cloud models, no subscription tools.
7. Custom Development – All components custom-built. No SaaS, licensed, or commercial tools.
8. Termination – If milestone fails after revisions, client may cancel with no penalty.
9. Delivery – Milestones delivered in phases, but final must be one integrated system.
10. Payment & Retention – 50% per milestone released on approval, 50% held until final integrated platform delivered & tested.
11. Risk Management – If developer fails final delivery, client keeps retained balances + delivered code/assets.

*Clarification – Risk Management
Main goal: deliver the complete integrated platform — not isolated or low-quality tools.
The only win-win is achieving the fully connected platform at required quality.
This is why 50% hold-back applies per milestone.

---

*Screening Questions (Answer ALL)
1. Do you work as a team?
2. Do you agree with all stated conditions?
3. Can you deliver the platform fully offline with all 18 functions at the required quality?
4. Do you have AI experience (image, voice, animation, lip-sync)? Which tools are you expert in?
5. Why should I choose you over others?
6. Can you show samples proving you are qualified?
7. Can your team also build the Hasaballa Engine in the future (first major update after 6 months)?

Please provide a detailed work plan. Quality is the key — reference videos are the minimum acceptable level.

*Important
- Reference videos = minimum acceptable quality.
- Quality is non-negotiable.
- Project only awarded to a team with practical proof + clear plan.

*Only serious teams with proven AI experience should apply." (client-provided description)