Designed chatbot as a foundation for BI/search use-cases (KPIs, knowledge assistants) Developed prototype with NLP, and LangChain SequentialChains
Implemented contextual memory and RAG pipelines with FAISS embeddings
- AI Developer
- Back-End Developer
AI systems builder using Airtable, n8n, Langchain, and various more integration layers and complementary tools. Experience with building full-stack operations systems using Airtable.
AI systems builder using Airtable, n8n, Langchain, and various more integration layers and complementary tools. Experience with building full-stack operations systems using Airtable.
Designed chatbot as a foundation for BI/search use-cases (KPIs, knowledge assistants) Developed prototype with NLP, and LangChain SequentialChains
Implemented contextual memory and RAG pipelines with FAISS embeddings
Shipped an internal client-onboarding dashboard (Lovable + Next.js, deployed on Railway) with embedded Tavus AI live-avatar streaming, transforming onboarding from static intake to a conversational flow. Architected a German-market Meta Ads operation in Airtable with a 5-table model, 5 automations, ad-copy template library, and client-facing/internal interfaces wired to a Kanban approval flow (Zur Prüfung → Freigegeben → Veröffentlicht). Co-led client POC sessions and owned day-to-day client communication across the engagement.
Delivered 2+ production-grade multi-agent client lead qualification systems inside n8n using sub-agents, HTTP requests, and sub-workflow execution, improving qualification efficiency by 60%. Produced RAG SOP guides for standardized document ingestion using a self-hosted LightRAG and automated document upload/ingestion pipeline in n8n (JSONL). Enhanced agent development efficiency by 65% with prompt-generation/optimization frameworks. Engineered concise, jurisdiction-aware system prompts with tool behavior, safety guardrails, privacy, and hijacking resistance while collaborating with senior leadership across discovery sessions.
Built an AI-native cultural platform with an agent-based architecture using Claude Code, Cursor, and GPT-o3, scaling to support 345+ Saudi artists. Increased traffic through organic social funnels and GA4 monitoring. Used prompt-driven development with AI-generated code, automated debugging, and fallback systems. Engineered backend using Google Sheets as CMS with real-time API transformations and serverless functions. Built stakeholder/developer handover SOPs for continuity and build–test–deploy governance.
Purdue University & SimpliLearn
OpenAI & Deeplearning.ai
Microsoft
LangChain & Deeplearning.ai
Post a job and invite Ayman, or contact them directly. No hidden commission.