FeaturedOpportunity Radar20,000+ opportunities tracked worldwide, filtered for your expertise.One-time report — $29 $9.90 →

TwinKnowledge

AI agents that answer questions and flag conflicts across construction documents

AI platform that indexes an AEC firm's contracts, drawings, specifications, RFIs, and BIM models so teams can query project knowledge instantly and validate submittals against requirements. Uses a computer-vision pipeline to parse drawing sets and a construction-domain-tuned LLM to answer questions and flag scope conflicts before construction begins.

$3.7M Series Seed (April 2025) led by Camber Creek, with Great Wave Ventures raisedEarly Stage
Founded 2023 · Named pilot/case-study customers include U.S. Space Force, Toll Brothers, SHoP Architects, and Sound Transit; total distinct customer count not publicly disclosed
Website →← All Tools

Why This Tool Exists

The Problem

AEC teams store project knowledge across scattered contracts, drawing sets, specifications, RFIs, and BIM models, making it slow to answer questions or verify compliance. The vendor cites industry figures that average RFI response times run 6.4-10 days and that misaligned contract documents drive the majority of downstream project problems.

The Solution

TwinKnowledge connects to a firm's document repositories and design tools, then lets users query project information through an AI assistant and create custom agents scoped to a project or company. The system validates submittals and designs against contracts, specifications, and internal checklists, and surfaces scope conflicts before they cause rework. Published case studies (vendor-reported) describe SHoP Architects reducing time spent on RFI responses from 75% to 25% of CA staff time, and Sound Transit increasing as-built drawing review throughput roughly 13x.

How You Use It

Delivery Method
SaaS
Integrations
AutoCADBluebeamRevitProcore
Project Phases
Design Development, Construction Documentation, Construction Administration
Project Types
Commercial, Residential, Institutional, Infrastructure

Data Transparency

Exactly what this tool uses and how

Input
What it needs
Required:Construction documents (contracts, specifications, RFIs), Drawing sets
Optional:BIM models, Internal standards and QA/QC checklists, Building codes
Formats:PDF, DWG
Output
What you get
Format:Conversational AI answers and validation reports surfaced in a web assistant
Fields:Document-grounded answers with citations, Flagged scope conflicts, Compliance validation results, RFI answer suggestions
Algorithm
How it works
Model:Proprietary computer vision model (fine-tuned per client/project) for parsing drawing sets, paired with a large language model fine-tuned on construction-domain language for document Q&A and retrieval (per a published AWS case study)
Accuracy:Vendor claims "100% coverage" analysis of drawing sets versus industry-standard spot-checking; case-study results (SHoP Architects, Sound Transit) are vendor-reported and not independently verified
Privacy
How your data is protected
Retention:Not publicly specified
Training:Not publicly specified whether customer data is used to train underlying models; CV models are described as fine-tuned per client and project
Compliance:SOC 2 Type II attestation claimed (annual; full report available on request via the vendor site), TLS 1.2+ encryption in transit; encryption at rest stated on vendor site
API
Integration
Endpoint:Contact vendor for API access
Method:Not publicly documented

Use Cases

  • ·Instant Q&A across contracts, RFIs, memos, and other project documents
  • ·Searching drawing sets for spec notes and details without manual page-by-page review
  • ·Searching BIM models for design components, assemblies, and products
  • ·Validating designs and submittals against internal standards and building codes
  • ·Flagging scope conflicts between contracts, drawings, and specifications before construction begins

Pricing

Free
Not publicly specified - no free tier advertised
Pro
Contact vendor for pricing
Enterprise
Custom pricing for enterprise AEC firms and public agencies
Website →
Research Sources & Data QualityLast verified: 2026-07-16
Verified Data (9)
problem, solution, deliveryMethod, integrations, yearFounded, fundingTotal, algorithm-approach, privacy-compliance-claims, customerBase-named-logos
Not Found (6)
pricing-details, api-documentation, data-retention-period, ai-training-data-use, exact-customer-count, team-size
Our Commitment: We only include verified data from official sources. If information isn't publicly available, we mark it as "Not publicly specified" rather than guessing.