SAP Business One & AI Integration Engineer (MCPs, TecDoc, WMS & Shopify)

We are looking for a senior, hands-on **AI Systems Architect & Integration Engineer** to serve as our long-term technical partner over the next **12 months** (part-time, **~10 hours per week**). You will lead the technical design, backend orchestration, and deployment of a corporate **AI Operating System ("AI OS" / Central Brain)** for a major enterprise B2B distributor (220+ employees, 100k+ SKUs). The company is moving away from isolated chatbots and fragmented spreadsheets. Our goal is to build a centralized, secure data and AI context layer that connects our ERP, PIM, WMS, demand engines, and frontline field communication (WhatsApp/Portals), enabling automated business skills, agentic workflows, and real-time decision intelligence. #### What You Will Build & Own * **The "AI OS" Middleware:** Design and maintain the central data ingestion, retrieval (RAG), and context engine connecting to internal databases and external platforms. * **ERP & Systems Integration:** Build bidirectional integrations, event-driven pipelines, and custom connectors across a multi-system enterprise stack. * **Custom MCP Servers & Tool Calling:** Develop and deploy Model Context Protocol (MCP) servers and Function Calling endpoints so LLM agents can query live inventory, invoice aging, and catalog compatibility safely. * **Automated Operational Skills:** Deploy production workflows for frontline operations, including: * Automated morning route briefings for 24 field sales reps (overdue balances, stock availability, buying patterns) via WhatsApp/Portal. * Automated payment reconciliation and invoice matching rules. * Daily executive intelligence briefs. * **Enterprise Governance & Security:** Implement Role-Based Access Control (RBAC) and privacy guardrails ensuring sensitive financial/payroll data is partitioned and protected within a private cloud environment (AWS most likely). --- #### Required Technical Stack & Experience We are looking for proven experience in the following specific environment: 1. **ERP & Enterprise Databases:** * **SAP Business One:** Extensive experience with the SAP B1 **Service Layer (REST / OData)**, DI API, and writing optimized SQL/HANA queries. * Relational databases (PostgreSQL, MS SQL Server, Supabase). * Experience working with legacy/proprietary database architectures (GeneXus C# web portals/WMS, custom PIMs) is a major plus. 2. **AI & Agentic Frameworks:** * **Model Context Protocol (MCP):** Real-world experience building or configuring custom MCP servers to bridge LLMs with closed APIs and internal databases. * **LLM Orchestration:** Advanced prompt engineering, Function Calling, and agentic workflows using **Claude (Anthropic API)** and **OpenAI API**. * RAG pipelines, vector databases (pgvector/Pinecone/Qdrant), and hybrid keyword/semantic search. 3. **Integration, Middleware & APIs:** * **Automation Engines:** Heavy hands-on experience with **n8n** (self-hosted) or **Make**. * **Backend:** Python (FastAPI/Flask) or Node.js/TypeScript for microservices, webhooks, and middleware connectors. * **Messaging & Frontline Channels:** Meta WhatsApp Cloud API / WhatsApp Business Platform (implementing Human-in-the-Loop workflows). * Standard REST APIs (TecAlliance/TecDoc, Streamline, E-commerce APIs). 4. **Data Analytics:** Familiarity building data pipelines feeding into **Power BI** or **Looker Studio**. --- #### Working Structure & Commitment * **12-Month Project:** Structured roll-out starting with infrastructure and core quick wins, followed by phased operational skills across 4+ business divisions. * **Part-Time Engagement:** Expected dedication of **~10-15 hours per week**. * **Weekly Strategic & Technical Syncs:** One mandatory 60-minute weekly meeting with our leadership team (conducted in Spanish or English) to review completed sprint tickets, test deployed skills, and prioritize upcoming milestones. * **Long-Term Mindset:** We are not looking for a one-off script writer. We want an engineering thought partner who will actively evolve the company's AI architecture over the year. --- #### How to Apply (Mandatory Requirements) In your proposal, please provide: 1. **Loom Video Intro (3 to 5 minutes – required):** Introduce yourself, summarize your direct experience with enterprise ERP integrations (especially SAP Business One) and MCP/agent development, and briefly describe how you would approach structuring a centralized "AI OS" for an organization with 10+ disjointed software systems. 2. **Past Case Study/Repo:** A link, repo, or architecture diagram illustrating a complex workflow, agent, or data pipeline you engineered in production. *(Proposals without the requested Loom video will not be reviewed).*