AI Tendering Agent for Construction
Budget: £5,000 – £10,000 GBP
Title
AI engineer / ML engineer (freelance) — Build AI agent for tendering, take‑offs and cost estimating (construction)
Overview
We are a small residential building company seeking an experienced AI/ML freelancer or boutique team to build an AI agent that automates our tendering process. The agent must: read drawings and site survey photos, perform accurate material and labour take‑offs, cost items using our actual supplier prices, and output complete tender documents ready for client submission.
Key Tasks:
- Preparing bid documents
- Take-offs from drawings
- costing BoQ
Potential Formats:
• CAD / technical drawings
• DWG, DXF (AutoCAD)
• DGN (MicroStation)
• Vector PDF (architectural/engineering drawings)
• SVG
• BIM / 3D model formats
• IFC (industry standard for BIM)
• RVT (Revit)
• OBJ, FBX, STL (mesh formats)
• Point clouds / survey data
• LAS / LAZ, E57, PLY
Images & photo evidence
• JPEG / JPG, PNG, HEIC, TIFF (incl. multi‑page TIFF)
• RAW (Canon/ Nikon/ Sony etc.)
• Ensure ability to read EXIF metadata and geotags
Scans / documents
• Searchable/OCR-friendly PDF (PDF/A where possible)
• Multi-page PDFs, TIFFs
• Text & office formats
• DOCX, RTF, TXT
• XLSX, CSV (spreadsheets; price lists, labour rates, BOQ exports)
• Excel templates used by your ERP
• Web / API data
• REST APIs (JSON), SOAP/XML, HTML scraping of supplier portals
Email / messaging
• EML / MBOX (for parsing email attachments/requests) — helpful but optional
• Geospatial / mapping (if applicable)
• GeoJSON, Shapefile (.shp)
Miscellaneous
• Annotated/overlay formats (DXF layers, PDF markups)
• Versioned files and common archive formats (ZIP, 7z)
Ideal Skills & Experience
Technical ML / CV
• Strong ML/computer‑vision building production CV systems (object detection, segmentation, layout analysis).
• Experience with document understanding (OCR, layout parsing, table extraction, semantic labeling).
• Familiarity with transfer learning, few‑shot learning, and strategies for limited labelled data.
• Experience training/validating models and producing performance metrics (precision/recall, IoU, error vs. ground truth).
CAD / BIM / drawing parsing
• Parsing and extracting geometry from DWG/DXF, vector PDF, IFC/RVT, and converting between formats (RVT→IFC, raster→vector).
• Comfortable using CAD/BIM libraries/tools (e.g., Teigha/ODA, ezdxf, Autodesk Forge, IfcOpenShell, pyRevit).
Image & photogrammetry
• Photo analysis skills: perspective correction, plane detection, material/condition classification from mobile photos.
• Experience with point clouds / LAS/E57 optional but valuable.
Backend, integration & devops
• Backend/API development (Python preferred; frameworks like FastAPI/Django).
• Familiar with ML infra: model serving (TorchServe, TensorFlow Serving, or containerised endpoints), CI/CD, Docker, Kubernetes or serverless deployments.
• Cloud experience (AWS/Azure/GCP) for training, storage and secure hosting.
• Experience integrating with REST APIs, web scraping, and handling supplier portals.
Data engineering & pipelines
• ETL/data‑pipeline experience: ingesting PDFs/DWGs/photos, normalising formats, annotation pipelines, data versioning.
• Familiar with label tools and managing annotation quality.
Product & UX
• Ability to design simple review/edit UI for estimators (annotated drawings, override edits) or to integrate with existing tools.
• Deliver audit trails and traceability from quantities to source drawings/photos.
Security, compliance & governance
• GDPR awareness and secure handling of client/site images and personal data.
• Experience with secure credentials handling, encryption at rest/in transit, and role-based access control.
Testing, validation & maintenance
• Experience creating test suites, acceptance criteria, and model monitoring/drift detection.
• Plan for retraining lifecycle and maintenance SLAs.
Domain knowledge (highly desirable)
• Construction estimating / quantity surveying experience, or prior projects with BOQ/tender outputs.
• Familiarity with UK construction conventions, measurement rules (NRM/SMM), common material/labour breakdowns.
I am looking for a freelance engineer (or small specialist team) to build an end‑to‑end AI agent that automates take‑offs, costs tenders from drawings and site photos using my actual supplier prices, and integrates with supplier/accounting systems and my labour rates to produce a comprehensive tender document fully costed to submit to clients
AI engineer / ML engineer (freelance) — Build AI agent for tendering, take‑offs and cost estimating (construction)
Overview
We are a small residential building company seeking an experienced AI/ML freelancer or boutique team to build an AI agent that automates our tendering process. The agent must: read drawings and site survey photos, perform accurate material and labour take‑offs, cost items using our actual supplier prices, and output complete tender documents ready for client submission.
Key Tasks:
- Preparing bid documents
- Take-offs from drawings
- costing BoQ
Potential Formats:
• CAD / technical drawings
• DWG, DXF (AutoCAD)
• DGN (MicroStation)
• Vector PDF (architectural/engineering drawings)
• SVG
• BIM / 3D model formats
• IFC (industry standard for BIM)
• RVT (Revit)
• OBJ, FBX, STL (mesh formats)
• Point clouds / survey data
• LAS / LAZ, E57, PLY
Images & photo evidence
• JPEG / JPG, PNG, HEIC, TIFF (incl. multi‑page TIFF)
• RAW (Canon/ Nikon/ Sony etc.)
• Ensure ability to read EXIF metadata and geotags
Scans / documents
• Searchable/OCR-friendly PDF (PDF/A where possible)
• Multi-page PDFs, TIFFs
• Text & office formats
• DOCX, RTF, TXT
• XLSX, CSV (spreadsheets; price lists, labour rates, BOQ exports)
• Excel templates used by your ERP
• Web / API data
• REST APIs (JSON), SOAP/XML, HTML scraping of supplier portals
Email / messaging
• EML / MBOX (for parsing email attachments/requests) — helpful but optional
• Geospatial / mapping (if applicable)
• GeoJSON, Shapefile (.shp)
Miscellaneous
• Annotated/overlay formats (DXF layers, PDF markups)
• Versioned files and common archive formats (ZIP, 7z)
Ideal Skills & Experience
Technical ML / CV
• Strong ML/computer‑vision building production CV systems (object detection, segmentation, layout analysis).
• Experience with document understanding (OCR, layout parsing, table extraction, semantic labeling).
• Familiarity with transfer learning, few‑shot learning, and strategies for limited labelled data.
• Experience training/validating models and producing performance metrics (precision/recall, IoU, error vs. ground truth).
CAD / BIM / drawing parsing
• Parsing and extracting geometry from DWG/DXF, vector PDF, IFC/RVT, and converting between formats (RVT→IFC, raster→vector).
• Comfortable using CAD/BIM libraries/tools (e.g., Teigha/ODA, ezdxf, Autodesk Forge, IfcOpenShell, pyRevit).
Image & photogrammetry
• Photo analysis skills: perspective correction, plane detection, material/condition classification from mobile photos.
• Experience with point clouds / LAS/E57 optional but valuable.
Backend, integration & devops
• Backend/API development (Python preferred; frameworks like FastAPI/Django).
• Familiar with ML infra: model serving (TorchServe, TensorFlow Serving, or containerised endpoints), CI/CD, Docker, Kubernetes or serverless deployments.
• Cloud experience (AWS/Azure/GCP) for training, storage and secure hosting.
• Experience integrating with REST APIs, web scraping, and handling supplier portals.
Data engineering & pipelines
• ETL/data‑pipeline experience: ingesting PDFs/DWGs/photos, normalising formats, annotation pipelines, data versioning.
• Familiar with label tools and managing annotation quality.
Product & UX
• Ability to design simple review/edit UI for estimators (annotated drawings, override edits) or to integrate with existing tools.
• Deliver audit trails and traceability from quantities to source drawings/photos.
Security, compliance & governance
• GDPR awareness and secure handling of client/site images and personal data.
• Experience with secure credentials handling, encryption at rest/in transit, and role-based access control.
Testing, validation & maintenance
• Experience creating test suites, acceptance criteria, and model monitoring/drift detection.
• Plan for retraining lifecycle and maintenance SLAs.
Domain knowledge (highly desirable)
• Construction estimating / quantity surveying experience, or prior projects with BOQ/tender outputs.
• Familiarity with UK construction conventions, measurement rules (NRM/SMM), common material/labour breakdowns.
I am looking for a freelance engineer (or small specialist team) to build an end‑to‑end AI agent that automates take‑offs, costs tenders from drawings and site photos using my actual supplier prices, and integrates with supplier/accounting systems and my labour rates to produce a comprehensive tender document fully costed to submit to clients