AI Developer Needed: Build a Plumbing & Mechanical Estimating Tool (Claude 3.5 Sonnet / Vision LLM)

We are a plumbing and mechanical contracting company looking to build a specialized AI-assisted estimating tool. The goal of this tool is to help our estimating team review construction documentation, perform visual spot-checks on mechanical blueprints, and analyze commercial construction contracts (specifically CCDC documents and supplementary conditions). We want to leverage Claude 3.5 Sonnet (via Anthropic API or Claude Projects) due to its superior performance with complex layout analysis and vision tasks. Key Functions the Tool Needs to Perform:Contract Risk Analysis: Scan heavy PDF contract files (CCDC 2, CCDC 5B, and client Supplementary Conditions) to instantly flag strict notice periods, harsh indemnification clauses, and liquid damages. Visual Document Spot-Checking: Allow estimators to feed in mechanical/plumbing drawings (floor layouts, riser diagrams, mechanical room details) to accurately parse and extract info from equipment schedules, valve charts, and text callouts on the blueprints. Scope & Spec Verification: Cross-reference written project specifications with drawing callouts to highlight discrepancies or missed scope items before we bid. Technical Requirements for the Developer: Proven experience working with Vision LLMs (specifically Claude 3.5 Sonnet or GPT-4o) for layout-aware document processing. Experience with advanced PDF parsing and preprocessing (handling large-format blueprint PDFs, multi-column text, and vector graphics).Deep understanding of RAG (Retrieval-Augmented Generation) frameworks to ensure the AI queries massive document sets without hallucinating numbers. Experience with Python or node.js, and integrating AI outputs into structured formats (JSON, or direct export to Excel/CSV templates).Bonus: Prior experience building tools for the construction, AEC (Architecture, Engineering, Construction), or manufacturing industries.