Osama A.

⭐ Top Rated | 100% Job Success | Senior AI Engineer @ NexAI (Riyadh) | Research Assistant | Kaggle Expert 🇪🇬 Most AI projects fail somewhere between the demo and production. I close that gap. I'm Osama Senior AI Engineer currently leading a team at NexAI building enterprise AI products. Over the past few years I've shipped 10+ production systems for startups and enterprise clients: not prototypes that crumble under real load, but tools people inside the company actually use every day. My focus is agentic systems and RAG pipelines — the category where most freelancers can demo something impressive and deliver something disappointing. If you've been burned before, I get it. Here's what's different: you see a working prototype within days, every deliverable ships with documentation your team can maintain, and I'll tell you when an idea won't work before you pay for it to fail. ──────────────────────────────────────── RECENT WORK ──────────────────────────────────────── 🤖 Multi-Agent Stock Analysis System Four specialized agents in sequence: real-time news fetching → sentiment scoring → technical indicator analysis → trading signal output. Built to production-grade multi-agent architecture standards. Full code on GitHub. Stack: LangGraph · CrewAI · LangChain 📚 Production RAG Platform (SaaS client) Ingests 1,000+ documents into Weaviate with hybrid search, conversation memory, and a full FastAPI + Next.js interface. Delivered in under two weeks. The client had real customers using it before the month ended. Stack: Weaviate · LlamaIndex · FastAPI · Next.js 🏛️ InstaMasr AI Assistant for Egyptian Startup End-to-end: AI chat assistant, intelligent form automation, Arabic RTL support, JWT auth, dark/light theming. Architecture to production deployment, solo. Stack: OpenAI · FastAPI · PostgreSQL · React 🔄 Lead Capture Automation WhatsApp → n8n → CRM → Slack. Lead comes in, full onboarding sequence triggers automatically. Cut two hours of daily manual work to zero. Stack: n8n · Make · Custom webhooks ──────────────────────────────────────── FULL TECH STACK ──────────────────────────────────────── AI & LLMs: LangChain · LangGraph · LlamaIndex · CrewAI · GPT-4 · Hugging Face · LoRA/QLoRA fine-tuning Vector DBs: Weaviate · Pinecone · Milvus · ChromaDB · FAISS · pgvector ML/Data: PyTorch · TensorFlow · Scikit-learn · OpenCV · Pandas · NumPy Backend: FastAPI · Next.js · React · TypeScript · PostgreSQL · Docker Cloud: AWS · GCP · Railway · Vercel Automation: n8n · Make · Custom APIs ──────────────────────────────────────── WHY CLIENTS COME BACK ──────────────────────────────────────── → You see working code fast. No weeks of "discovery" before I show you something real. → Clean, documented code. Whoever maintains this after me won't be lost. → Kaggle Expert + HackerRank Top 98th percentile in Egypt (Python). I compete, not just copy. → 100% Job Success. Every project delivered. → I'll tell you when an idea won't work before you pay to find out. ──────────────────────────────────────── HOW WE START ──────────────────────────────────────── 1. You describe the problem — even roughly. "I want AI to handle X" is enough. 2. I give you an honest assessment: what's possible, best approach, realistic timeline. 3. Working prototype within days. 4. We refine together, then ship with documentation. ──────────────────────────────────────── LET'S TALK ──────────────────────────────────────── Send a message. I don't take every project — if it's not the right fit, I'll tell you and point you somewhere better. If it is, we move fast. I typically reply within a few hours. 📌 GitHub: github/Osama-Abo-Bakr