NbAIl – The Next-Gen Multimodal AI Assistant
Problem Statement 1 – Building on Groq API
Problem Statement 5 – Build Next-Gen Command Agents (Terminator)
Today's AI assistants are too slow, too basic, or too disconnected from real user needs.
NbAIl solves this by combining super-fast text, smooth voice conversation, intelligent command execution, and planned augmented reality (AR) capabilities — all in a single multimodal powerhouse.
One of NbAIl’s proudest innovations: Command Mode.
Simply type a command like /open notepad and write hello world, and our Terminator Agent, powered by Groq for understanding and local Python automation, will open system apps, type messages, and perform actions automatically.
Zero manual work. Full smart automation.
Built to behave like a true futuristic assistant — not just a chatbot.
NabilThange
- Nabil (Lead Developer / Designer / Researcher)
- Prioritized multimodal interaction — text, voice, and AR — from the start.
- Focused heavily on ultra-fast response speed, natural voice support, and full command automation.
- Treated UI/UX design as a first-class citizen — not an afterthought.
- Built a modular system ready for expansion into more complex real-world use cases.
- Frontend: Next.js, Tailwind CSS, Framer Motion
- Backend: Groq LPU Cloud API
- Voice: Vapi Voice API
- Command Execution: Custom-built Terminator Python Agent
- Hosting: Vercel for Frontend, Ngrok for Local Tunneling
- ✅ Groq: Lightning-fast LLM for conversation + command parsing.
- ✅ Screenpipe: Planned for future AR mode integration.
- ✅ Blazing Fast AI Chat (text-based conversations powered by Groq)
- ✅ Voice Assistant Mode (ask, command, and get instant voice responses)
- ✅ Command Mode (/open appname and act) — dynamic, real-time command execution inside apps
- ✅ Augmented Reality (AR) Mode — screen-aware AI (prototype)
- ✅ World-Class UI/UX — smooth animations, intuitive layout, mobile-first responsive design
- ✅ Multi-mode Switcher — instantly switch between chat, voice, command, and AR
- ✅ Error Handling & Fail-Safes — graceful recovery when apps not found
🚀 NbAIl doesn’t just "chat." It "acts."
- Demo Video Link: https://youtu.be/oY1w7WNAXSA?si=9_MI7qxENo5M-_gT
- ✅ All members followed 2+ official social channels and filled the form
- ✅ Bonus Task 1 completed (Badge Sharing) — 2 Points
- ✅ Bonus Task 2 completed (Sprint.dev signup) — 3 Points
- Node.js (v18+ recommended)
- Python 3.10+ (for Terminator Agent)
- Groq API Key
- Vapi Voice API Key
- Ngrok Account (for tunneling local Terminator Agent)
.envfile setup for API keys
# Clone the repo
git clone https://github.com/your-team/nbail
# Install frontend dependencies
cd nbail
npm install
# Start the frontend
npm run dev
# ⚡ How to Run Terminator Agent
### Step 1: Open a New Terminal Window
- Navigate to the `terminator-agent` folder.
### Step 2: Install Python Dependencies
```bash
pip install -r requirements.txt
### Step 3: Run the Terminator Agent
```bash
python terminator_agent.py
### Step 4: Start an Ngrok Tunnel
```bash
ngrok http 8000
(Replace 5000 with your local port if it’s different.)
—
###✅ 4. important-notes
- Terminator Agent runs locally for full security and direct access to your apps.
- Command Feature is not available to the public because:
- There are no free Windows VM providers.
- Cloud VM providers like AWS, Azure require credit/debit cards, and I am not old enough to have one yet.
- HTTPS is required for using voice and AR modes — Ngrok provides secure HTTPS tunnels for your local server.
- Terminator Agent only works on your machine — making it faster, more private, and customizable!
Note: Always make sure Terminator Agent and Ngrok are running before using command features inside NbAIl!
While NbAIl already delivers a futuristic multimodal experience, the journey has just begun!
Here’s what’s planned for future versions:
- Execute system commands and app actions using pure voice input.
- Example:
"Hey NbAIl, open Word and write a letter."
- Using Screenpipe and real-time OCR models, NbAIl will understand your live screen content.
- Contextual actions based on what's visible — like suggesting to save a document if you're working in Word.
- Chain multiple commands together intelligently.
- Example:
/open Chrome -> search for HackHazards site -> take screenshot
- Enable conversations and commands in multiple languages:
Hindi, Spanish, Japanese, French, and more!
- Allow users to build and share plugins for new commands.
- Extend NbAIl with custom actions (e.g., open Photoshop and resize images automatically).
- Encrypt all voice, text, and command data using secure protocols.
- Offer local-only processing options for users who want maximum privacy.
- NbAIl can enter a special Productivity Mode where it blocks distractions (e.g., closes games, opens study apps) automatically.
- Bring NbAIl to mobile devices and wearable AR glasses for anywhere, anytime assistance.
This project would not have been possible without the amazing technologies, APIs, and open-source tools available to the community.
Here’s what powered NbAIl:
- Next.js - Frontend framework
- Tailwind CSS - Beautiful, responsive UI styling
- Three.js - 3D visuals on website
- Groq AI - Ultra-fast LLMs for real-time interaction
- Screenpipe: Terminator - For Doing System Level work on command
- Vapi Voice API - Stores Voice Cache
- Ngrok - Local development tunneling for HTTPS -Spline - 3d Interactive Model
- Visual Studio Code - Code editor
- Figma - UI/UX designing
- Vercel - Hosting frontend (for demos)
- Cursor - Backend and Bug Fixing and Final Touches
- J.A.R.V.I.S. from Iron Man (fictional AI assistant)
- Midjourney + DALL-E for creative UI references
- Hackathons like HackHazards25 for pushing limits!
Building NbAIl has been an incredible journey — full of challenges, learning, and excitement.
What started as just an idea quickly turned into a living, breathing system that combines
the power of AI, system control, and futuristic user experience.
- AI isn't just about answering questions — it's about understanding context and acting on it.
- A clean, fast, and intuitive UI/UX is just as important as powerful features.
- Thinking big and executing small steps every day brings even the craziest ideas to life.
- NbAIl is not just a project; it's a vision of the future — where digital assistants truly understand and help users at every moment.
- Even with limitations like no access to cloud VMs (due to age restrictions), I found creative ways (local hosting + ngrok) to bring the vision alive.
- Every error, every crash, every late night spent debugging — it all made the final build even more special.
Bigger, smarter, and even crazier updates are on the way.
Thank you for reading, testing, and believing. The future is ours to create.
"The ones who are crazy enough to think they can change the world are the ones who do."
— Apple, 1997
