Platform Agnostic Interface to Assistant

Job Request: Multi-AI Assistant Chat Platform (Project-Based UI with Memory & Image Integration)
Objective:Build a sleek, platform-agnostic AI assistant interface that supports modular connections to multiple LLM providers (OpenAI, Grok, Qwen, DeepSeek). The system should emulate modern chat UIs (like ChatGPT) while enabling project-based memory, time-filtered conversation history, and optional tools like a visual canvas and image generation integration.
🧠 Core Features
1. Unified Streaming Chat UI
Realtime assistant conversation flow with streaming output
Support for switching or routing queries to different AI providers (OpenAI, Grok, Qwen, DeepSeek) via selectable dropdown or auto-routing
2. Persistent Memory + Project Organization
Conversations stored by project folders or time windows (Today, Yesterday, Last 7 Days, Older)
Option to tag, rename, or archive threads
Support for user-editable memory snippets, tied to AI assistant personas or platform contexts
3. Modular Multi-LLM Backend
API connectors for:
OpenAI (Assistants API + GPT models)
Grok by xAI (when API access available)
Qwen (Alibaba Cloud)
DeepSeek Chat API
Ability to toggle or prioritize LLM routing per session or per message
4. Rich Text + Canvas Mode
Switch into a collaborative Canvas/TextDoc mode (for longform editing, outlining, brainstorming)
Save and return to canvas files within project folders
5. Visual Tools Integration
Plug in image generation platforms (DALL·E, Midjourney, OpenArt, etc.) via API
Send/receive prompt-based image generations inline with text threads
6. Clean, Minimal UX
Think: Notion x ChatGPT x Figma Canvas
Responsive design, optimized for focus and creative work
🛠 Tech Stack Preferences
Frontend: React (preferred), Svelte, or Vue
Backend: Node.js, Python (FastAPI or Flask), or comparable modern stack
Real-time features: WebSockets or SSE for chat streaming
Authentication (optional): OAuth or JWT; user-specific memory management
🎯 Deliverables
Fully functional web-based prototype
Working connections to at least 2 AI APIs (OpenAI + one other)
Project/memory structure with file saving (text or markdown)
Canvas-style editing for longform content
Visual design that feels elegant, light, distraction-free
💬 When Applying, Please Include:
Brief background or links to past relevant work
Suggested approach for modular LLM connectivity
Preferred tech stack and reasoning
Rough timeline and estimated hours/cost
Your take on building clean, distraction-free UX for assistant workflows
 ... Show more