Lead AI Systems Architect & Automation Head

Job TitleLead AI Systems Architect & Full-Stack Automation Head (Salary in USD)

Job DescriptionThe Mission:We are a highly profitable ($1.2M/year profit)

US-based firm.

We don’t hire for "tasks"—we hire Intelligence.

We are looking for a Founding Engineer in Egypt to act as the sole architect of an ever-evolving ecosystem of AI-driven products.

You will have a direct impact on reducing overhead and maximizing revenue through pure technical innovation.

If you want to build the "Brain" of a multi-million dollar company, this is your seat.The Capabilities You Must Possess:We are looking for the top 0.1% of engineers in Egypt.

You must be a "Generalist Specialist" capable of rapid prototyping and deployment across any domain we dream up: Autonomous Intelligence: Mastery of LLM Orchestration (GPT-4o/Claude), Vector Databases (Pinecone/Milvus), and RAG.

You build systems that learn.

Global Voice Infrastructure: Expertise in VoIP/SIP (Asterisk/Kamailio) and Real-time Audio (

TECHNICAL STACK REQUIREMENTS

1.

The "Scraper & Lead Gen" Pillar (Data Acquisition)

Advanced Browser Automation: Expert in Playwright or Puppeteer (Stealth-plugins are mandatory).

Anti-Bot Evasion: Proven ability to bypass Cloudflare Turnstile, DataDome, and Akamai using JA3 fingerprinting, TLS header manipulation, and residential proxy rotation (Bright Data/Oxylabs).

Asynchronous Processing: High-concurrency scraping using Python Asyncio or Go to pull thousands of leads without bottlenecking.

2.

The "Voice & Dialing" Pillar (Infrastructure)

VoIP/SIP Engineering: Deep experience with Asterisk, FreePBX, or Kamailio.

Ability to manage SIP Trunks and DID (Phone Number) pools.

Regulatory Compliance (US/Global): Knowledge of STIR/SHAKEN protocols and "A-Level Attestation" to prevent calls from being labeled "Scam Likely." Low-Latency Audio: Expertise in WebRTC and Vapi/Retell/Telnyx to keep AI voice response time under 500ms.

• Server Management: Ability to deploy local or cloud-based "Gateway Servers" in specific regions to bypass regional AI dialing restrictions.

3.

The "Training Brain" & RAG Pillar (Intelligence)

Vector Databases: Implementation of Pinecone, Milvus, or Weaviate to store and retrieve successful sales call data.

Multimodal RAG: Building systems that don't just "read" text but "analyze" audio files to identify winning sales patterns from top reps.

Speech-to-Speech (S2S): Integration with GPT-4o Audio or ElevenLabs to create human-like voices with emotional intelligence (tone, laughter, pauses).

Diarization & Transcription: Using OpenAI Whisper to split recordings into "Rep" vs. "Customer" for automated performance grading.

4.

The "Automation & CRM" Pillar (Logic)

Agentic Frameworks: Proficiency in LangGraph, CrewAI, or AutoGen to build "Agents" that can autonomously update the CRM, book appointments, and handle customer service.

Airtable/CRM Mastery: Deep API integration with Airtable, HubSpot, or GoHighLevel using Webhooks and Python SDKs.

Mobile App Integration: Ability to build a lightweight Flutter or React Native app interface for D2D guys to access the "Real-time Training Brain" in the field.

TECHNICAL STACK REQUIREMENTS

1.

The "Scraper & Lead Gen" Pillar (Data Acquisition)

Advanced Browser Automation: Expert in Playwright or Puppeteer (Stealth-plugins are mandatory).

Anti-Bot Evasion: Proven ability to bypass Cloudflare Turnstile, DataDome, and Akamai using JA3 fingerprinting, TLS header manipulation, and residential proxy rotation (Bright Data/Oxylabs).

Asynchronous Processing: High-concurrency scraping using Python Asyncio or Go to pull thousands of leads without bottlenecking.

2.

The "Voice & Dialing" Pillar (Infrastructure)

VoIP/SIP Engineering: Deep experience with Asterisk, FreePBX, or Kamailio.

Ability to manage SIP Trunks and DID (Phone Number) pools.

Regulatory Compliance (US/Global): Knowledge of STIR/SHAKEN protocols and "A-Level Attestation" to prevent calls from being labeled "Scam Likely." Low-Latency Audio: Expertise in WebRTC and Vapi/Retell/Telnyx to keep AI voice response time under 500ms.

• Server Management: Ability to deploy local or cloud-based "Gateway Servers" in specific regions to bypass regional AI dialing restrictions.

3.

The "Training Brain" & RAG Pillar (Intelligence)

Vector Databases: Implementation of Pinecone, Milvus, or Weaviate to store and retrieve successful sales call data.

Multimodal RAG: Building systems that don't just "read" text but "analyze" audio files to identify winning sales patterns from top reps.

Speech-to-Speech (S2S): Integration with GPT-4o Audio or ElevenLabs to create human-like voices with emotional intelligence (tone, laughter, pauses).

Diarization & Transcription: Using OpenAI Whisper to split recordings into "Rep" vs. "Customer" for automated performance grading.

4.

The "Automation & CRM" Pillar (Logic)

Agentic Frameworks: Proficiency in LangGraph, CrewAI, or AutoGen to build "Agents" that can autonomously update the CRM, book appointments, and handle customer service.

Airtable/CRM Mastery: Deep API integration with Airtable, HubSpot, or GoHighLevel using Webhooks and Python SDKs.

Mobile App Integration: Ability to build a lightweight Flutter or React Native app interface for D2D guys to access the "Real-time Training Brain" in the field.