Raidium

Name: Raidium
Brand: Raidium

AI Native

The AGI radiology company Between Paris & Palo Alto [StartX F2025]

Est 2022|Valuation $55.6M|Raised $18.2M

LinkedIn 21K|Web traffic 2K|FTEs 390|HQ Palo Alto, California, US

Published Jul 14, 2026

01Common complaints

1. Organizations that require mature, externally validated enterprise proof (extensive hospital references, abundant third-party reviews) prior to purchasing; public references for this vendor are limited. 2. Organizations that require immediately available, documented security/compliance artifacts (SOC 2 Type II, ISO 27001 certificates, pen-test summaries, detailed security whitepapers) in vendor public materials. These artifacts were not located in public pages reviewed. 3. Sites where integration change must be minimal and there is limited capacity to support viewer replacement or substantial workflow change. 4. Organizations that prefer narrowly scoped, point-solution AI modules with well-established regulatory status for specific indications (e.g., stroke, PE). Those buyers may prefer established vendors offering specific cleared algorithms. Sources: https://www.raidium.eu/ ; https://www.raidium.eu/legal-notices.html ; https://axisimagingnews.com/radiology-products/radiology-software/pacs/raidium-unveils-new-ai-native-pacs-viewer

Disqualifiers (poor fit)

Organizations requiring extensive third-party validation, numerous public references, and mature compliance attestations prior to purchase
Sites with constrained IT integration capacity or limited resources for validation and change management
Use cases requiring immediately available, widely regulated point-solution modules with extensive real-world references

Vendor-acknowledged limitations

No explicit public "known issues" page or detailed roadmap with dates was located in the reviewed materials.

Practical limitations and evidence gaps

- No public pricing, which complicates ROI forecasting and vendor comparisons. - Security and compliance posture is not clearly documented in public materials (no SOC 2/ISO 27001 references found). - Limited accessible third-party user reviews and case studies, making validation of usability, support quality, and real-world performance difficult. - Regulatory maturity appears to be in progress, based on job postings that reference QMS and regulatory submissions. Sources: https://www.raidium.eu/ ; https://www.welcometothejungle.com/fr/companies/raidium/jobs/qara-apprentice_paris

02Why they buy

Vendor-claimed features (as presented by the company and press)

- An integrated viewer approach with an embedded foundation model intended to provide consolidated interaction capabilities rather than relying solely on bolt-on AI tools. - The Curia foundation model, described by the vendor as multimodal and multitask for 3D imaging, with vendor statements about training dataset scale and broad task coverage. - Prompt-driven workflow execution (text and visual prompts) with the radiologist maintained in the decision loop, as described in vendor materials. Note: these are vendor-claimed features and should be validated through technical documentation, demonstrations, and independent evaluation. Sources: https://www.raidium.eu/ ; https://raidium.eu/agi-research/ ; https://axisimagingnews.com/radiology-products/radiology-software/pacs/raidium-unveils-new-ai-native-pacs-viewer

Where competitors may have advantages

Incumbent PACS vendors: enterprise hardening, integration depth, procurement familiarity, and large reference bases.
AI module vendors: narrower scope but often clearer regulatory posture and established clinical performance per indication.
AI orchestration platforms: may enable adoption without changing the primary viewer, depending on deployment architecture.

03Field reports

Review platforms (G2 / Gartner Peer Insights / TrustRadius)

- Gartner Peer Insights: searches did not return an accessible Raidium review page; Gartner content is often access-restricted and early-stage vendors may have limited coverage. - TrustRadius: no Raidium review page was identified in public searches. - G2: the vendor has a G2 product listing and an alternatives page, but substantive user review text was not available in the accessible pages. Sources: Gartner Peer Insights public site; G2 product and category pages; TrustRadius public listings

Community (Reddit / Hacker News)

A Hacker News thread references the Curia arXiv preprint. Direct access to the Hacker News discussion in the browsing attempt encountered an access error, so quoted community commentary could not be retrieved. Reddit searches returned primarily job postings mentioning Raidium rather than practitioner reviews.

Sources: Hacker News discussion pages; Reddit community threads and postings

Vendor-provided endorsements and labeling

The vendor homepage lists organizations under a labeled section (the vendor uses the label "Trusted by innovators"). These are endorsement signals listed by the vendor; they are not equivalent to independently verified customer references or detailed case-study outcomes.

Sources: Official vendor website and marketing pages

Conclusion on customer sentiment

Credible, detailed end-user reviews were not publicly available in the sources reviewed. There is insufficient public evidence to characterize user sentiment (support quality, bug frequency, ROI, and usability) with high confidence. This represents a limitation for procurement teams that rely on third-party validation.

Public case studies

No formal, detailed public case studies with quantified operational ROI (e.g., time saved per read, throughput impact, diagnostic performance in routine practice) were found on the vendor site during this research.

Research publications as evidence (distinction from case studies)

The vendor highlights research outputs such as ONCOPILOT (npj Precision Oncology) and a CURIA preprint on arXiv, plus conference abstracts. Research publications can indicate technical progress but do not replace operational case studies demonstrating integration outcomes, clinician adoption, or regulatory deployment. Critical note: where performance claims are presented, procurement teams should request access to study methodology, datasets, reader study details, and any external validation reports. Sources: Official vendor research and publications page

04Best fit

Organization sizes

Most plausible ICP (inferred from product type and funding stage): mid-market to enterprise imaging organizations (multi-site imaging groups, academic medical centers, oncology centers) that can fund integration, validation, and change management.

Industries / settings

Radiology departments
Oncology imaging workflows (vendor materials emphasize oncology focus)
Research hospitals and academic centers (biomarker discovery and clinical trials)

Sources: Official vendor research page; Public funding program portfolio listing

Roles and personas

Decision makers: CMIO, Radiology Chair, Imaging Informatics leadership, PACS administration, Procurement, IT security, Compliance/QARA
End users: staff radiologists, fellows/residents, oncology imaging specialists

Qualification criteria (good fit)

Organizations with the internal capacity for integration and clinical governance
Groups seeking to evaluate consolidated workflows that incorporate model-driven prompts and automation
Sites willing to engage in early-adopter or co-development relationships with the vendor

05The pitch

What the product does (objective core value proposition)

Raidium describes its product as an AI-integrated radiology viewer / PACS viewer intended to consolidate radiologists' workflows by embedding a foundation model (Curia) into the viewing experience rather than relying exclusively on multiple disconnected AI point solutions. The vendor presents the following capabilities: - A single viewer intended to perform multiple radiology tasks (segmentation, measurements, comparisons, reporting) via text and visual prompting. - A multimodal, multitask foundation model (Curia) described by the vendor as supporting clinical use and research-oriented imaging biomarker extraction. Sources: Official vendor product and research pages; industry press coverage announcing the product

Target market and positioning

Primary target users: radiologists, radiology departments, and imaging centers. Positioning: the vendor describes the product as an "AI operating layer" or unified interface for radiology intended to reduce workflow fragmentation associated with multiple task-specific AI tools and viewer switching. The company also describes ambitions toward broad automated reasoning in imaging (language such as "general intelligence in radiology" and "radiology AGI" appears in vendor communications); these claims are aspirational and not independently validated in public materials. Sources: Official vendor product and research pages; industry press coverage

Get a comprehensive analysis of Raidium including market position, competitive landscape, adoption trends, and peer benchmarks.

Download full report →

Published Jul 14, 2026

01Common complaints

Organizations that require mature, externally validated enterprise proof (extensive hospital references, abundant third-party reviews) prior to purchasing; public references for this vendor are limited.
Organizations that require immediately available, documented security/compliance artifacts (SOC 2 Type II, ISO 27001 certificates, pen-test summaries, detailed security whitepapers) in vendor public materials. These artifacts were not located in public pages reviewed.
Sites where integration change must be minimal and there is limited capacity to support viewer replacement or substantial workflow change.
Organizations that prefer narrowly scoped, point-solution AI modules with well-established regulatory status for specific indications (e.g., stroke, PE). Those buyers may prefer established vendors offering specific cleared algorithms.

Sources: https://www.raidium.eu/ ; https://www.raidium.eu/legal-notices.html ; https://axisimagingnews.com/radiology-products/radiology-software/pacs/raidium-unveils-new-ai-native-pacs-viewer

Disqualifiers (poor fit)

Organizations requiring extensive third-party validation, numerous public references, and mature compliance attestations prior to purchase
Sites with constrained IT integration capacity or limited resources for validation and change management
Use cases requiring immediately available, widely regulated point-solution modules with extensive real-world references

Vendor-acknowledged limitations

No explicit public "known issues" page or detailed roadmap with dates was located in the reviewed materials.

Practical limitations and evidence gaps

No public pricing, which complicates ROI forecasting and vendor comparisons.
Security and compliance posture is not clearly documented in public materials (no SOC 2/ISO 27001 references found).
Limited accessible third-party user reviews and case studies, making validation of usability, support quality, and real-world performance difficult.
Regulatory maturity appears to be in progress, based on job postings that reference QMS and regulatory submissions.

Sources: https://www.raidium.eu/ ; https://www.welcometothejungle.com/fr/companies/raidium/jobs/qara-apprentice_paris

02Why they buy

Vendor-claimed features (as presented by the company and press)

An integrated viewer approach with an embedded foundation model intended to provide consolidated interaction capabilities rather than relying solely on bolt-on AI tools.
The Curia foundation model, described by the vendor as multimodal and multitask for 3D imaging, with vendor statements about training dataset scale and broad task coverage.
Prompt-driven workflow execution (text and visual prompts) with the radiologist maintained in the decision loop, as described in vendor materials.

Note: these are vendor-claimed features and should be validated through technical documentation, demonstrations, and independent evaluation.

Sources: https://www.raidium.eu/ ; https://raidium.eu/agi-research/ ; https://axisimagingnews.com/radiology-products/radiology-software/pacs/raidium-unveils-new-ai-native-pacs-viewer

Where competitors may have advantages

Incumbent PACS vendors: enterprise hardening, integration depth, procurement familiarity, and large reference bases.
AI module vendors: narrower scope but often clearer regulatory posture and established clinical performance per indication.
AI orchestration platforms: may enable adoption without changing the primary viewer, depending on deployment architecture.

03Field reports

Review platforms (G2 / Gartner Peer Insights / TrustRadius)

Gartner Peer Insights: searches did not return an accessible Raidium review page; Gartner content is often access-restricted and early-stage vendors may have limited coverage.
TrustRadius: no Raidium review page was identified in public searches.
G2: the vendor has a G2 product listing and an alternatives page, but substantive user review text was not available in the accessible pages.

Sources: Gartner Peer Insights public site; G2 product and category pages; TrustRadius public listings

Community (Reddit / Hacker News)

Sources: Hacker News discussion pages; Reddit community threads and postings

Vendor-provided endorsements and labeling

Sources: Official vendor website and marketing pages

Conclusion on customer sentiment

Public case studies

Research publications as evidence (distinction from case studies)

Critical note: where performance claims are presented, procurement teams should request access to study methodology, datasets, reader study details, and any external validation reports.

Sources: Official vendor research and publications page

04Best fit

Organization sizes

Industries / settings

Radiology departments
Oncology imaging workflows (vendor materials emphasize oncology focus)
Research hospitals and academic centers (biomarker discovery and clinical trials)

Sources: Official vendor research page; Public funding program portfolio listing

Roles and personas

Decision makers: CMIO, Radiology Chair, Imaging Informatics leadership, PACS administration, Procurement, IT security, Compliance/QARA
End users: staff radiologists, fellows/residents, oncology imaging specialists

Qualification criteria (good fit)

Organizations with the internal capacity for integration and clinical governance
Groups seeking to evaluate consolidated workflows that incorporate model-driven prompts and automation
Sites willing to engage in early-adopter or co-development relationships with the vendor

05The pitch

What the product does (objective core value proposition)

A single viewer intended to perform multiple radiology tasks (segmentation, measurements, comparisons, reporting) via text and visual prompting.
A multimodal, multitask foundation model (Curia) described by the vendor as supporting clinical use and research-oriented imaging biomarker extraction.

Sources: Official vendor product and research pages; industry press coverage announcing the product

Target market and positioning

Primary target users: radiologists, radiology departments, and imaging centers.

Positioning: the vendor describes the product as an "AI operating layer" or unified interface for radiology intended to reduce workflow fragmentation associated with multiple task-specific AI tools and viewer switching. The company also describes ambitions toward broad automated reasoning in imaging (language such as "general intelligence in radiology" and "radiology AGI" appears in vendor communications); these claims are aspirational and not independently validated in public materials.

Sources: Official vendor product and research pages; industry press coverage

Get a comprehensive analysis of Raidium including market position, competitive landscape, adoption trends, and peer benchmarks.

Download full report →