CADDi
by CADDi Inc.
AI data platform that structures fragmented manufacturing drawings, specs and purchase history
CADDi is an AI data platform for manufacturers that ingests engineering drawings, CAD files, specifications, ERP records and procurement history and structures them into one searchable, agent-readable layer. It is built for procurement, design and quality teams who cannot find what the company already designed, already bought, or already got wrong — and therefore re-buy near-identical parts from different suppliers.
CADDi is a Tokyo- and Chicago-headquartered AI data platform for manufacturing, founded in 2017 and led by co-founder and CEO Yushiro Kato. Its original product, CADDi Drawer, attacked a specific and expensive problem: manufacturers repeatedly buy near-identical parts from different suppliers at different prices because nobody can find what the company already designed or procured. Drawer ingested technical drawings and searched a customer's own archives for similar or identical parts using keyword, similarity and image search. Over the two years to September 2026 CADDi broadened this into what it calls an AI data platform for manufacturing, built in four layers: a data layer that ingests and structures engineering and business data, a semantic layer that connects fragmented records across systems, a discovery-and-agents layer, and purpose-built workflow products. Drawer has been renamed CADDi Explorer and is now joined by CADDi Agent, which handles part-standardisation decisions and quality impact assessments, plus six workflow products — Composer, Design Review, Process Review, Cost Review, Quote and ALM — that capture the tacit judgement of experienced engineers as reusable data. Kato is explicit that general-purpose LLMs are used for documents while proprietary models handle technical files, on the grounds that LLMs do not understand drawings or CAD. More than half of Japan's 100 largest manufacturers use CADDi, deployed across 22 countries, with named customers including Sumitomo Machinery America, Sugino Machine and ASTEMO and published outcomes such as 100,000 hours saved at an automotive OEM. On 16 September 2026 CADDi raised a $114 million Series D at a $1.2 billion valuation — more than double its March 2025 mark — taking total funding to $234 million.
The VP of Procurement or Head of Engineering at a discrete manufacturer whose drawings, specs, quotes and defect history live in separate systems, so the same part gets redesigned and re-sourced instead of reused.
One searchable, agent-readable layer over drawings, CAD, ERP and purchase history that surfaces the part you already own — one industrial machinery customer reports a 30% annual procurement cost reduction.
At a Glance
- Category
- Industry & Government
- Pricing
- Contact for pricing, Subscription
- Target Market
- CIOs, CTOs, VP Procurement, Manufacturing Engineers, Quality Leaders
- Deployment
- Cloud-only
- Founded
- 2017
- Headquarters
- Tokyo, Japan and Chicago, IL, USA
- Team Size
- 500+
- Customers
- Used by more than 50% of Japan's 100 largest manufacturers; deployed across 22 countries
Key Features
- ✓CADDi Explorer (formerly Drawer)
Discovery engine over drawings and manufacturing data with keyword, similarity and image search across mixed legacy formats.
- ✓CADDi Agent
AI agent that analyses connected manufacturing data to support part-standardisation decisions and quality impact assessments autonomously.
- ✓Semantic layer across systems
Connects fragmented records spanning CAD, ERP, procurement and HR systems into one contextualised model for people and agents.
- ✓Six domain workflow products
Composer, Design Review, Process Review, Cost Review, Quote and ALM capture engineering decisions as reusable structured data.
- ✓Purchase-history linkage
Ties each drawing to what was actually bought, at what price, from whom, exposing duplicate sourcing and price variance directly.
- ✓Proprietary models for technical files
Uses in-house AI for drawings and CAD rather than general LLMs, which the CEO argues cannot read engineering files reliably.
Capabilities
Use Cases
- •Duplicate part consolidation
Procurement finds identical parts already sourced elsewhere in the group; one industrial machinery customer cut annual procurement cost 30%.
- •Regulated substance investigation
A semiconductor equipment manufacturer reports cutting investigation time by more than 90% by searching the structured drawing archive.
- •AI-assisted design review
CADDi Design Review flags likely errors in new drawings based on problems recorded against historically similar parts.
- •Engineering knowledge retention
Workflow products capture how experienced engineers make cost and process calls, preserving judgement as staff retire.
- •Global engineering data standardisation
A multinational OEM unifies drawings and specs across 22 countries of operations, reporting 100,000 hours saved.
Ideal For
Best For
- ✓Discrete manufacturers consolidating duplicate or near-identical part spend across plants and suppliers
- ✓Engineering teams that need visual and similarity search over decades of legacy drawings in mixed formats
- ✓Quality organisations running impact assessments and design reviews against historical defect history
- ✓Regulated-substance and compliance investigations across large drawing archives — one semiconductor equipment maker reports a 90%+ time reduction
- ✓Multinational manufacturers standardising engineering data across many country sites and ERP instances
Not Ideal For
- ✗Process manufacturers, software firms or any business without an engineering-drawing and BOM estate — the platform's value is specific to discrete manufacturing artifacts
- ✗Teams that need deep bidirectional API integration into custom internal systems: reviewers on Software Advice and Capterra repeatedly cite missing API integration and reliance on unstable workarounds
- ✗Buyers who want to try before committing — there is no free trial and no published price, so evaluation requires a vendor-led sales cycle
- ✗Organisations unwilling to fund change management, which CEO Yushiro Kato names as the primary obstacle to adoption
Deployment
Market & Ratings
Used by more than 50% of Japan's 100 largest manufacturers; deployed across 22 countries
Market Analysis
Pros
- ✓Very strong verified user sentiment — 4.8/5 across 43 reviews on Capterra and Software Advice, with ease of use 4.9 and customer support 4.9 and no reviews below 3 stars
- ✓Reviewers consistently praise search quality specifically: accurate keyword search, good similar-drawing matching, fast response, and visual rather than purely textual retrieval
- ✓Credible enterprise security posture for a vertical vendor: ISO/IEC 27001, SOC 2 Type 2, ITAR access control, GDPR, SSO/MFA, encryption in transit and at rest, and regular third-party assessments
- ✓Deep market proof rather than logo slides — more than half of Japan's 100 largest manufacturers, 22 countries, sales more than doubling year over year, and a $1.2B Series D valuation
Cons
- ✗Integration is the single most consistent complaint in reviews: users report the service 'lacks the ability to integrate with our functions like APIs', forcing unstable workarounds that hurt cost-effectiveness
- ✗Reviewers describe quotes, orders and drawing search as disconnected tools rather than one flow, and ask for better data-table readability and finer viewing permissions
- ✗Change management is the CEO's own stated primary adoption obstacle — the platform only pays back if engineers actually change how they search and source
- ✗The review base is small and Japan-weighted (43 reviews), so the 4.8 rating is a real but thin signal for a North American or European buyer
- ✗No public pricing and no free trial, and the product surface has expanded fast — Drawer was renamed Explorer and six workflow products plus CADDi Agent launched recently, so maturity varies module to module
Pricing
Platform (quoted)
Contact for pricing
- ✓CADDi Explorer discovery engine
- ✓CADDi Agent
- ✓Workflow products
- ✓Customer success onboarding
CADDi publishes no list pricing: Capterra records the starting price as 'Contact vendor' and explicitly notes that no free trial is available, so every deal is quoted. The company staffs more than 100 customer success employees out of roughly 900 total headcount, which signals a high-touch, implementation-heavy enterprise motion rather than seat-based self-serve — budget for onboarding effort alongside licence cost.
Security & Compliance
Sources
This page was written from 6 sources, 4 on domains other than caddi.com.
Stay Ahead of the Curve
Weekly enterprise AI insights for technology leaders. No spam, no vendor pitches—unsubscribe anytime.
SubscribeRelated Products
Faro Health
Agentic AI that turns clinical trial protocols from static documents into structured development intent
Lumafield Quality Agent
Physical-AI quality agent that catches unanticipated manufacturing defects and traces root causes from industrial CT and sensor data
Sixfold
An AI underwriter that learns your book and takes submissions through to bind-ready
Casepoint IQ
Governed agentic AI for eDiscovery, FOIA and investigations, on FedRAMP High and DoD IL6 infrastructure