Daniel Francisco L.Private profile: Upwork shows its earnings only to signed-in members, so we check it every few weeks, not daily. MRR here is a rough estimate (Upwork's six-month earnings ÷ months active). Last read Nov 7, 2026.

LangGraph and CrewAI multi-agent systems, RAG pipelines grounded in your own data, and n8n automations that run unattended. If your AI prototype works in testing but falls apart with real users, that's exactly what I fix. ⚙️ WHAT I DELIVER 🤖 Multi-Agent Systems LangGraph, CrewAI, AutoGen. Agents with memory, tool-calling, self-correctionloops and guardrails that handle non-linear workflows instead of one-shot prompts. 📚 Production RAG Hybrid search, re-ranking, GraphRAG, semantic caching. Grounded in your documents, built to resist hallucination, with evaluation so you can prove accuracy instead of hoping for it. 🛠️ AI SaaS & Backend Python, FastAPI, streaming (SSE/WebSockets), vector DBs (Pinecone, Qdrant, pgvector, Chroma), plus TypeScript/Next.js when the frontend needs to match. 🔗 Workflow Automation n8n, custom API integrations, and internal tools that remove manual steps from operations, support, and research. 🧠 WHY CLIENTS KEEP ME ON I architect systems, I don't wrap APIs. My custom agent framework (Nexus) was built from scratch stateful reasoning loops, agent memory, blast-radius analysis, observability. That background means I catch the failure modes that cause AI projects to stall at 80% done. 📦 Recent builds: 🔐 A-SOC — multi-agent security operations system using agent consensus for real-time threat triage 🧪 CodeGuardian — automated code review agent with reasoning traces 📊 OpsManager — agentic workflow layer replacing manual operational handoffs 🔎 Research Intelligence — multi-source RAG with source attribution 💡 I also tell you when you don't need agents. If a single well-designed prompt chain or an n8n workflow solves it for a tenth of the budget, I'll say so before you spend the money. 🤝 HOW WE'LL WORK ✅ Clear scope before we start ✅ Progress you can see, not status reports ✅ Communication in business outcomes, not framework names ⚡ Usually reply within a few hours can scope most projects same-day 📩 Send me your use case, even a rough one, and I'll tell you honestly whether I'm the right fit. 🏷️ CORE STACK AI Agents | Multi-Agent Systems | Agentic Workflows | LangGraph | LangChain | CrewAI | AutoGen | LlamaIndex | DSPy | RAG | GraphRAG | Hybrid Search | Vector Databases | Pinecone | Qdrant | pgvector | Weaviate | ChromaDB | LLM Integration | OpenAI API | Anthropic API | Llama | Gemini | Fine-Tuning (LoRA/QLoRA) | Prompt Engineering | LLMOps | AI Chatbot Development | AI SaaS | Python | FastAPI | Django | TypeScript | Node.js | Next.js | React | REST API | WebSockets | n8n | Workflow Automation | Business Process Automation | AI Agent Developer | AI Architect