Build a Local AI Operations Platform for Procurement, RFQs, Logistics & Finance

Build a Local AI Operations Platform for RFQs, Procurement, Logistics, Finance & Email Automation We are looking for an experienced AI automation engineer / full-stack developer to design and build a secure local AI Operations Platform for a UAE-based defense and aviation company. This is not a chatbot project. The objective is to build an operational AI system that manages the complete workflow from receiving an RFQ or ten der until project delivery and financial closure. The platform will initially support Defendex Partners and later Defenix Logistics. ⸻ Core Business Workflow The system should support the following end-to-end workflow: RFQ Received → Document Processing → Procur ement Assignment → Supplier Outreach → Quotation Collection → Commercial Comparison → Bid Submission → Purchase Order → Logistics Execution → Finance Follow-up → Project Closure ⸻ Main Modules 1. RFQ & Tender Processing The system should: * Monitor designated folders and/or email for new RFQs. * Read PDF, Excel, Word and scanned documents. * Perfor m OCR on scanned files. * Extract: * RFQ Number * Platform / Aircraft * Part Numbers * Description * Quantity * UOM * Condition * Deadlines * Technical Notes led * On Hold ⸻ 1. Procurement* Generate standardized Excel files. * Create procurement tasks. * Track RFQ status: * New * In Sourcing * Quoted * Submitted * Extended * Awarded * Lost * Cance l Management The platform should: * Assign RFQs to procurement staff. * Recommend suppliers based on platform, OEM and previous history. * Generate supplier RFQ emails. * Track supplier responses. * Send follow-up reminders. * Record: : * Compan * Price * Lead Time * Certifications * Condition * Currency * Payment Terms * Compare quotations. * Recommend the best commercial option. ⸻ 1. Supplier Intelligence Database Maintain supplier profiles includin gy Information * Contacts * Countries * Platforms Supported * OEM Relationships * Certifications * Repair Capability * Stock Capability * Lead Time History * Response Speed * Reliability Rating * Previous Transactions ⸻ 1. Email Operations Agent * Prepare draft replies. * Support appIntegrate with Microsoft 365 / Outlook. The system should: * Read incoming emails. * Classify emails. * Detect: * New RFQs * Extensions * Quotations * Purchase Orders * Shipment Updates * Invoices * Link emails to projects .roval workflows before sending emails. ⸻ 1. Commercial & Quotation Module The system should: * Compare supplier quotations. * Calculate: * Cost * Freight * Total Landed Cost * Selling Price * Margin * Gross Profit * Working Capital * Generate quotation comparison reports. ⸻ 1. Finance Module Track: * Customer Invoices * Supplier Invoices * Receivables * Payables * Cash Flow * Project Profitability * Outstanding Payments * Payment Due Dates * Weekly and Monthly Cash Flow Forecasts ⸻ 1. Logistics Module After Purchase Order: Track: * Shipment Status * AWB * Packing List * Commercial Invoice * Certificates * Customs Documents * Dangerous Goods Documents * Delivery Status * Proof of Delivery * Delays ⸻ 1. Employee Daily Reporting Collect structured reports from: * Procurement * Logistics * Finance * Operations Generate one consolidated daily management report. ⸻ 1. CEO Dashboard Display: * Active RFQs * Upcoming Deadlines * Quotations Received * Purchase Orders * Active Shipments * Delayed Shipments * Outstanding Payments * Cash Position * Expected Profit * Staff Activity * Operational Alerts ⸻ Preferred Technology Open to recommendations, but preferably: * Docker * Python * FastAPI * PostgreSQL * Vector Database (Qdrant / pgvector / Weaviate) * Ollama or another local LLM runtime * OCR Engine * REST APIs * Role-Based Access Control * Microsoft Graph API * React / Next.js Dashboard AnythingLLM may be used as a knowledge interface, but it should not be the core operational platform. ⸻ Security Requirements The platform will manage commercially sensitive defense and aviation information. Therefore: * Local deployment * Role-based permissions * Audit logs * Secure credential storage * Backup and recovery * No external sharing of confidential company data ⸻ Important Development & Delivery Requirements T his is a long-term project, but it will be developed in multiple phases. We do not expect the entire platform to be completed in one milestone. Instead: * The developer will build each module independently in their own development environment. * Development should initially use dummy/sample data. * Each completed module must be delivered as a complete deployment package. Each delivery must include: * Full source code * Docker configuration * Database migrations * Documentation * Installation guide * Configuration guide We will install and deploy every completed phase inside our own local environment before approving the next milestone. The developer should not require direct access to our production environment, company servers, company email accounts or confidential business data. Production credentials, email accounts and sensitive company information will only be configured by our internal team after delivery. The architecture must support incremental deployment so that each completed phase can be installed, tested and expanded without redesigning the entire platform. ⸻ Required Experience Please apply only if you have experience with: * AI Agents * Workflow Automation * OCR * Document Processing * Python * PostgreSQL * Vector Databases * Microsoft Graph API * Outlook Integration * Docker * Local LLMs * Enterprise Software * Procurement / Supply Chain systems (preferred) ⸻