👋 180degree is a leading AI consultancy helping brands across MENA and Europe deploy cutting-edge voice and automation solutions.
We’re building a production-grade Syrian Arabic AI Voice Agent for Charco Chicken, a popular Syrian restaurant chain expanding rapidly across Turkey.

This Hiring Quest is your opportunity to prove your skills, get hired full-time, and contribute to one of the region’s first production-ready, multilingual AI ordering agents – starting with real-time voice automation over phone calls.

 🕓 Start Date: Immediate
🌍 Location: Remote
💰 Salary: $600 to $2500

🛠️ How the Hiring Quest Works

1. Register for the quest

2. After the registration period ends, you’ll receive email the quest process details

3. Submit your solution before the deadline

4. Top candidates will be invited to a review session

5. One candidate will be hired, others may be shortlisted for future opportunities

🔍 Who We’re Looking For

1. 0–5 years of experience in AI/NLP/voice/speech projects

2. Strong Python skills (NLP pipelines, API integration, etc.)

3. Familiar with STT (Speech-to-Text) and TTS (Text-to-Speech) tools

4. Experience building Arabic or multilingual NLP models

5. Deep knowledge of voice persona design and dialogue flow

6. Experience with SIP trunking, VoIP, and telephony systems (required)

7. Bonus: Frontend/backend skills for dashboards or testing tools

🎯 Your Mission: Build a Syrian Arabic Voice Agent with Real-Time Telephony Support

🧩 Business Context

Charco Chicken wants to automate its phone order system using a voice agent that speaks flawless Syrian Arabic and can handle end-to-end customer orders. Your prototype should demonstrate:

1. Real-time SIP call handling

2. Natural, culturally relevant Syrian conversation

3. Backend API to simulate order processing

4. Easy-to-use UI for additional testing

🛠️ Your Tasks

📞 1. SIP Trunk Integration (Core Requirement)

1. You must enable real-time voice interactions via SIP trunk or cloud telephony provider (e.g., Asterisk, Twilio, Vonage)

2. Users should be able to call a phone number and speak to the voice agent

3. Handle voice I/O with low latency and high reliability

4. Include deployment and SIP config instructions in the ReadMe

⚠️ Critical Evaluation Criterion: Your agent must be callable through a SIP number and respond naturally in a flawless Syrian Arabic accent.
Solutions not meeting this will be disqualified, regardless of other quality factors.

🎙️ 2. Voice Agent Interaction Logic

1. Accept voice input in Syrian Arabic (live or simulated)

2. Use STT (e.g., Whisper, Google, Azure) to transcribe input

3. Detect intent (order, complaint, question, etc.)

4. Generate contextual Syrian Arabic responses using TTS (e.g., PlayHT, ElevenLabs)

5. Implement empathetic handling, upsells, and fallback scenarios

🔁 3. Backend API

1. POST /submit-order

   1. Accepts name and order list

   2. Returns mock order ID and ETA

   3. Stores submissions in-memory or JSON

🧪 4. UI Testing Tool 

1. Build an interface (Streamlit, Gradio, Flask UI) that allows:

   1. Audio upload or typed Arabic input

      1. Display of: Transcribed input , Detected intent/entities, Agent’s response (text + audio)
         

2. Must work locally (streamlit run app.py) or via a shared link (bonus)

📊 5. (Bonus) Monitoring Dashboard

1. Show: Conversation logs, Detected intents/entities, Replayable responses
   

2. Any frontend framework welcome (Streamlit, HTML, Flask, etc.)

🧰 Tech Stack

1. Python (Required)

2. STT/TTS: Whisper, ElevenLabs, Google Cloud, Azure, PlayHT

3. NLP: spaCy, transformers, Rasa, custom logic

4. Telephony: Twilio, Asterisk, SIP trunk services

5. Frontend: Streamlit, Flask, HTML

📦 What You Should Submit

🎥 Demo Video

1. Real-time call demonstration (via SIP or cloud telephony)

2. UI testing flow

3. Backend/API walkthrough

📁 GitHub Repository

Include:

1. Complete code

2. submit-order API

3. Setup instructions (including SIP config)

4. Tool choices explained

5. Scenario guide for how to test the agent

6. (Bonus) Deployed links to testing UI or dashboard

📊 Evaluation Criteria

1. 🎯 Syrian Arabic Accent Accuracy : 40% ✅ MUST be flawless – indistinguishable from a native speaker

2. 🤖 NLP Logic & Scenario Handling 25%

3. 🧠 Technical Implementation (SIP + STT/TTS + API) 20%

4. 📁 Code Quality & Documentation 15%

5. 🧪 Bonus: UI Testing Tool & Monitoring +10%

🚀 Bonus Points If You:

1. Use real Syrian Arabic TTS voices with strong prosody

2. Include empathy, fallback handling, upselling in agent flow

3. Add voice replay + intent summary to UI

4. Design an architecture that supports EN/TR multilingual extension
   

📩 After Submission

Shortlisted candidates will be invited to a live technical session to:

1. Walk through your agent and architecture

2. Review your design choices and edge-case handling

🕒 Final decision within 3–5 business days