🧠 Job Post: Build Full Offline AI Media Generation Platform – Hasaballa AI
⚠️ Please: Only serious teams with proven AI experience should apply.
🛠️ Platform: Hasaballa AI – Description & How it Works
The Hasaballa AI platform will run fully offline — even if it’s 200 GB in size.
🎯 Platform Goals
To produce unlimited videos at high speed / top quality for Arab creators — without any subscriptions, APIs, or internet connection — matching (at minimum) the quality shown in the following examples:
- https://tinyurl.com/4py6ntv5
- https://tinyurl.com/457h6pxb
- https://tinyurl.com/kye234sk
- https://tinyurl.com/s65thcdc
- https://tinyurl.com/mjb7cw8v
- https://tinyurl.com/r6b2ndn9
🧠 How to Use Hasaballa AI Platform
🟢 1. Smart Director – Automatic Mode (One Click – Full Auto Workflow)
- User writes the full script in Hasaballa GPT chat window (scenes, images, voices, animation...).
- Clicks the “Smart Director – Auto” button.
- System automatically generates everything based on the script:
• Image generation
• Character animation
• Voice generation & cloning
• Lip sync
• Background audio
• Final video timeline
🟡 2. Smart Director – Manual Mode (Full Hands-on Control)
- Script is written in Hasaballa GPT chat.
- Clicks “Smart Director – Manual.”
- System breaks the script into production stages step by step.
- User decides per stage:
• Generate image or skip
• Generate voice or skip
• Animate or not
• Add background or skip
• Activate lip sync or not
- Each element added to final timeline.
🔵 3. Standalone Mode (Use Each Tool Individually)
- Each tool (Image, Voice, Animation, Lip-Sync, Audio) has its own button.
- Tools can be used individually.
- Example: generate only images or test voices.
- ❌ Final video assembly is manual in this case.
⚠️ Platform Design Requirement – Hasaballa Engine Compatibility
The platform must be designed future-proof to integrate with the upcoming Hasaballa Engine (Task 19).
Hasaballa Engine will be an untrained core AI model that runs alongside all tools (Chat, Image Generation, Voice, Animation, Lip-Sync).
It will continuously learn by observing how these tools operate, the outputs they generate, and how the user creates videos.
Over time, the Engine should reach a level where it can replace 90–95% of the tools’ functionality, becoming the main backbone of the platform.
It will later be divided into sub-models (Text, ImageGen, Animator, Voice), which can also be offered online as paid APIs.
Therefore, all tools, pipelines, and logic must be modular and built in a way that allows seamless future integration with Hasaballa Engine — without requiring major redesigns.
---
📋 Hasaballa AI – 18 Core Components
1️⃣ System Requirements – Runs on Windows 11 (RTX 5090, 64GB RAM, 4TB SSD, WQXGA display). Must be modular and support future Hasaballa Engine integration.
2️⃣ Offline Execution – 100% offline: no APIs, no subscriptions, no credits and no internet dependency.
3️⃣ Arabic Interface – 100% Arabic RTL, single-page layout with all 18 integrated tools.
4️⃣ Chat Window (Hasaballa GPT) – Central hub to write full scripts, analyze scenarios, insert reference images/voices, and control Auto/Manual flow.
5️⃣ Image Generation – Ultra-realistic visuals (16:9, 9:16, 1:1), cinematic lighting, identity lock, text-to-image, image-to image Ai and Arabic text support.
6️⃣ Image Animation – Full-body + facial animation, lip sync, duration/camera control, optional cinematic effects (Please Test SkyReels V2).
7️⃣ Voice Cloning & Generation – Studio-grade voices, emotional tones, pauses/breathing, full Arabic dialect coverage (21 arabic dialectics) + multilingual support.
8️⃣ Background Audio – Built-in sound FX library (emotions, war, nature, crowd, etc.) + optional Pixabay search (Needs button to connect to the Internet).
9️⃣ Smart Director – Two modes:
• Auto Mode = full pipeline (script → final video).
• Manual Mode = step-by-step generation with user control.
-🕒 For a 7-minute video: total production time from click to final output must not exceed 45 minutes (includes ≈85 images, animation, lip sync, voice, sound, etc.).
-⏱️ For a 15-minute video: total time from click to final output must not exceed 90 minutes (≈170 images, full production)
🔟 🎞️ Timeline Editor & Final Video Editor (1–15 mins)
- Export options: HD, FHD, 4K (16:9, 9:16, 1:1).
- Full CapCut-style editor: trim/split, transitions, filters, effects, overlays, zoom/pan.
- Add external media (images, video, audio, text).
- Subtitles & captions: manual or AI (multi-language, sync, style, translation).
- Drag & drop timeline with placeholders for skipped steps (auto/manual Smart Director).
- Download, preview, and fully editable layered timeline.
1️⃣1️⃣ Smart Edit Layer – Edit video using Arabic prompts with timecodes (replace visuals, sounds, etc.).
1️⃣2️⃣ Integration with Google AdSense & YouTube 💻.(Needs buttons to connect to the Internet).
1️⃣3️⃣ Arabic Interface (Single Page Layout).
100% RTL support, 📌 Large buttons — each with a simple label and tooltip.
1️⃣4️⃣ Auto-Update Button 🔄.
(Requires internet) — used to update the platform in future if needed.
1️⃣5️⃣ Save Project & Continue Later 💾.
Automatically save the project every 30 seconds (only if changes occurred).
1️⃣6️⃣ Project Folder System – Local storage of all media assets per project.
1️⃣7️⃣ Backup & Restore – Full backup of models, settings, projects for safe recovery.
1️⃣8️⃣ Source Code Ownership – 100% exclusive rights to client, no reuse or resale allowed.---
👥 Preferred Team Composition
- AI/ML Engineers
- Computer Vision/Animation Specialists
- Backend Developers (Python, FFmpeg, CUDA)
- Frontend/Fullstack Developers (Arabic RTL UI)
- QA/Testers & System Optimizers
- Video Processing Experts
---
📂 Documentation
A full detailed document of the platform (including all tasks, sub-tasks, and requirements) will be provided to the selected developer/team upon award of the project.
💰 Budget: up to $7,500 (secured in escrow)
💎 Bonus: An additional $500–$1,000 will be paid upon successful delivery of the complete platform at the required quality standards.
📝 Mandatory Screening & Contract Conditions
All contract conditions and mandatory screening questions are clearly documented here:
Hasaballa AI Contract Conditions
👉 https://ahmedhasab.github.io/hasaballa-submissions/
⚠️ Important:
Every team that wishes to apply must visit the link above, answer all required questions, and submit the form.
Answers will be sent directly to me by email and will be the main basis for comparing proposals.
Applications without a completed submission will not be considered.
🔑 Payment Terms
The full budget ($7,500) will be deposited into escrow at the start of the contract.
Funds will only be released to the development team once the entire Hasaballa AI platform is delivered as one complete, fully integrated system at the required quality.
The purpose is to receive the full platform, not partial stages or separate tools.
⚖️ Quality Conditions
Reference videos represent the minimum acceptable quality.
Quality is non-negotiable.
The project will only be awarded to a team that provides:
1. Practical evidence of capabilities.
2. A clear, detailed work plan.
⚠️ Please: Only serious teams with proven AI experience should apply.
... Show more