AI Workflow Automation Developer

We are looking for a highly experienced AI Automation Engineer to design and build a sophisticated AI Agent ecosystem that automates the day-to-day operations of our engineering consulting and special inspections company. This is not a simple chatbot project. We are looking for someone who understands modern AI agent frameworks (OpenAI, LangGraph, CrewAI, AutoGen, MCP, etc.) and can build a secure business workflow automation platform that performs real work across multiple software systems. Our long-term vision is to build an AI Operations Manager capable of handling 80–90% of our administrative and project management tasks while maintaining human approval where required. Project Goal Build an AI Agent platform that acts like a virtual operations team capable of: Reading emails Understanding client requests Creating proposals Sending invoices Scheduling inspections Following up with clients Updating QuickBooks Completing engineering forms Assigning work to employees Managing inspection workflows Tracking project status Phase 1 Requirements 1. Email Intake AI Agent Monitor Gmail continuously. The AI should: Read incoming emails Categorize requests Detect new projects Detect inspection requests Detect proposal requests Detect payment notifications Detect scheduling requests Detect customer questions Automatically create internal tasks. 2. Proposal Agent Generate proposals automatically using predefined company templates. Examples: Special Inspection Proposal Structural Observation Proposal Concrete Testing Proposal Engineering Services Proposal Automatically: Calculate pricing Select predefined service SKUs Create proposal PDFs Email proposal to client Log proposal history 3. QuickBooks AI Agent Integrate with QuickBooks Online. Functions include: Create customers Create estimates Convert estimates to invoices Send invoices Apply received payments Verify payments Reconcile invoice status Generate statements Create recurring invoices Use predefined service SKUs 4. Inspection Scheduling Agent Automatically: Read inspection requests Determine inspection type Determine project location Estimate inspection duration Assign inspector Create calendar event Email inspector Email customer confirmation Update CRM 5. Employee Task Assignment Agent Assign work automatically. Examples: Inspector assignments Engineering calculations Report writing Proposal preparation Client follow-up Permit filing Material testing Integrations may include: Gmail Google Calendar CRM Slack (optional) Microsoft Teams (optional) 6. Follow-Up Agent Automatically monitor: Proposal expiration Outstanding invoices Missing documents Inspector reports Customer approvals Vendor responses Project deadlines Generate reminder emails automatically. 7. NYC DOB Agent Automate repetitive DOB work where technically and legally appropriate. Examples include: Complete TR1 forms Complete TR8 forms Populate DOB PDF forms Retrieve project information from DOB BIS Retrieve project information from DOB NOW Prepare filing documents Assist with Special Inspection sign-offs (with appropriate user authentication and approval) Validate filing information before submission Note: Any action involving authenticated government systems or legally binding submissions must include secure authentication and configurable human approval before final submission. 8. PDF Automation Agent Read and complete engineering documents. Examples: TR1 TR8 PW1 AI1 Engineering forms Client forms Extract information from: Email Existing PDFs DOB records Company database Generate completed forms automatically. 9. Payment Verification Agent Monitor: QuickBooks payments ACH notifications Credit card payments Automatically: Mark invoices paid Notify accounting Update project status 10. Inspection Completion Agent After inspection completion: Receive inspector report Verify required documents Generate invoice Send invoice Update QuickBooks Archive report Notify project manager Desired Technology Stack We are open to recommendations, but experience with the following is highly preferred: AI OpenAI API GPT-5.x LangGraph CrewAI AutoGen MCP (Model Context Protocol) AI Agent orchestration Backend Python FastAPI Node.js (optional) Automation n8n Make.com Zapier Temporal Celery Databases PostgreSQL Supabase Pinecone (optional) Redis Integrations Gmail API Google Calendar API QuickBooks Online API Google Drive API PDF processing libraries OCR Document AI Security Requirements Experience implementing: OAuth Google Workspace authentication QuickBooks authentication Secure credential management Role-based permissions Audit logs Human approval workflows for sensitive actions Encrypted storage of API keys Nice to Have Experience with: Construction industry software Engineering firms Project management systems CRM development PDF automation Government portals AI document extraction Computer Vision Voice AI AI assistants Deliverables Working AI multi-agent platform Source code Documentation Deployment instructions Database schema API documentation Admin dashboard Error logging Monitoring Testing To Apply Please include: Examples of AI agent systems you have built. Experience with LangGraph, CrewAI, AutoGen, MCP, or similar frameworks. Experience integrating with QuickBooks Online APIs. Experience with Gmail and Google Workspace APIs. Experience automating PDF generation and form filling. Experience building secure approval workflows for authenticated systems. GitHub portfolio or live demos. Your recommended technology stack for this project. Estimated timeline for an MVP and a production-ready version. Your availability for a long-term partnership. Budget Open to discussion based on experience and proposed architecture. This is expected to become a long-term project with continuous feature development and AI expansion. We are looking for a technical partner who can evolve this platform into an intelligent operations system for our engineering business.