Protex AI

Name: Protex AI
Brand: Protex AI
Rating: 4.8 (4 reviews)

AI Native

Protex AI is an AI video-analytics platform that connects to existing fixed-position CCTV/IP cameras via RTSP, detects predefined unsafe behaviors and conditions using on-site edge processing, and converts detections into event logs, dashboards, and workflows for EHS and operations teams.

Est 2021|Valuation $1.8M|Raised $54.3M|13 Ref Customers

LinkedIn 12K|Web traffic 1K|FTEs 97|HQ Dublin 2, County Dublin, IE

Published Oct 15, 2026

01Common complaints

- If you require transparent, self-serve pricing and rapid procurement - The vendor provides quote-based pricing and does not publish standardized self-serve tiers. - If you do not have adequate CCTV coverage or camera system accessibility - The product’s effectiveness depends on consistent camera views of risk areas; poor angles, lighting, or occlusion can undermine detection. - If you require a broad out-of-the-box safety detection library with minimal configuration - Vendors that emphasize large pre-built detection catalogs may reduce configuration effort, but buyers should validate real-world accuracy. - If workforce relations make video analytics politically infeasible - Even with anonymization options, some organizations will face strong resistance; in such cases consider non-video alternatives (proximity sensors, collision-avoidance systems, wearables) or process-based changes. - If your primary need is a full EHS suite (incidents, audits, training, SDS, compliance calendar) - This product is a video analytics layer and not a full EHS management system; a comprehensive EHS platform may better meet those requirements. - Public limitations acknowledged by the vendor - Camera streams are processed live and not stored in full, but event clips are retained by default unless videoless mode is enabled; this is a governance consideration rather than a bug. - Roadmap / future features - Public roadmap details are limited in the sources reviewed; buyers should request roadmap and release plans during procurement. - Scale limitations - Not publicly quantified. Practical constraints likely include edge appliance throughput, network capacity, and operational capacity to handle event volumes. - Security or compliance gaps - SOC 2 and other certifications are not clearly publicly documented on the vendor’s public pages reviewed here. - Data residency options are not clearly documented publicly and should be validated for regulated environments. - API limitations or integration constraints - The vendor references integrations and APIs in marketing materials, but detailed API documentation and supported integration patterns were not publicly accessible in the reviewed materials.

Deal disqualifiers: what makes a poor fit

Minimal camera coverage or cameras that cannot reliably provide required streaming and resolution/frame rates.
Strong workforce or union opposition to video-based monitoring without an ability to adopt anonymization modes or alternative technologies.
Teams that lack capacity to respond to increased volumes of near-miss or behavior event data.

02Why they buy

- Documented privacy-related configuration options - Vendor materials describe edge processing with configurable anonymization modes (facial blur, full-body blur, ghosting, and a videoless mode) and state a non-biometric identification posture in public documentation. - Hybrid architecture described in public materials - Live streams are processed on-site with only short event clips captured on detection and stored encrypted in the cloud, according to vendor documentation. - Evidence from limited customer reviews - Available reviews include statements of positive onboarding experience and perceived product stability; these are limited in number and should be validated in broader customer reference checks. - Where competitors may have advantages - Competitors may provide broader pre-built detection libraries, additional hardware options, or larger public review footprints; buyers should validate fit against specific operational requirements.

03Field reports

- Availability of third-party customer reviews - G2 - G2 lists a limited number of reviews and a favorable average rating, but the sample size is not sufficient for statistically robust conclusions. - AWS Marketplace - AWS Marketplace mirrors external reviews in some listings. - Gartner Peer Insights / TrustRadius - No accessible product page with a significant volume of reviews was located during this review session. - Direct quotes (positive and negative) from accessible sources - Positive (G2 review, republished on AWS Marketplace) - "Working with Protex AI has been the best vendor interaction I've had in my career." (G2 review republished on AWS Marketplace.) - "Onboarding new sites, including camera integration and rule configuration, is remarkably intuitive, and we have experienced virtually no bugs." (Same review context.) - Positive (G2 review, republished on AWS Marketplace) - "I enjoy using Protex AI because there's not much additional hardware needed since we can leverage existing security cameras." (G2 review republished on AWS Marketplace.) - Negative / friction (G2 review, republished on AWS Marketplace) - "The only thing that's been slightly complicated is the camera selection process… Going through the camera views can be time-consuming… if one work area has a very large number of camera views." (G2 review republished on AWS Marketplace.) - Common themes: why customers chose Protex AI - Reuse of existing CCTV and reduced incremental hardware footprint (mentioned in multiple reviews). - Dashboards, heatmaps, and vehicle-pedestrian interaction mapping used as decision support. - Positive vendor responsiveness and onboarding experience in available anecdotal reviews. - Why some customers left for competitors (churn reasons) - Public review volume is small; there is not enough accessible data to identify consistent churn reasons. - Specific outcomes customers achieved (with metrics when available) - From a G2 review (republished on AWS Marketplace) describing outcomes in a customer context - One reviewer reports reductions in at-risk event rates and ergonomic injuries; these are reviewer statements and should be validated in diligence. - From vendor case studies - Vendor-published case studies include claims of incident reductions over specific time windows; these should be treated as vendor-reported outcomes and validated for definitions and methodology. - Customer pain points and frustrations mentioned in reviews - Camera selection effort for large camera estates; cost-effectiveness may require choosing subsets of cameras, which can slow multi-site rollout. - Deal-breakers and reasons customers didn't purchase - Public information on lost deals is limited; procurement teams should capture and analyze vendor-specific disqualifiers during procurement conversations.

Vendor-published case studies (available titles/claims)

Marks & Spencer
- Vendor case-study title reports a substantial reduction in incidents during an initial deployment window.
Major UK packaging manufacturer
- Vendor case-study title reports reductions in safety events and references funding-related outcomes.
Bendix Commercial Vehicle Systems
- A case study exists but detailed metrics were not examined in this review session.

Critical evaluation of case studies

- Verification caveat - Case study metrics are published on vendor-controlled pages and are not independently audited in the accessible sources reviewed here. - Methodology gaps to look for (questions to pose to the vendor during diligence) - Definitions: what counts as an "incident," "safety event," or "at-risk event rate"? - Baseline window selection and seasonality adjustments. - Potential confounding interventions (for example, training campaigns, layout changes, staffing changes). - Data completeness: camera coverage and whether metrics cover all shifts/sites.

04Best fit

Company sizes

Best fit appears to be multi-site operators and large single-site facilities with meaningful incident exposure and existing CCTV coverage.
Practical range (inference)
- Upper mid-market to enterprise: organizations with a substantial employee base, multi-facility operations, and formal EHS programs.

Industries and verticals where adoption is strongest (with customer examples)

Retail and warehouse distribution
- Vendor materials reference a major UK retailer in case studies and external coverage.
Industrial manufacturing
- Vendor materials reference industrial customers and case studies.
Food distribution and manufacturing
- Case studies and vendor content reference food and chemical manufacturers.

Roles and personas: decision makers AND end users

Decision makers
- VP/Director of EHS, Head of Safety, VP Operations, plant and facility leadership, and risk or insurance stakeholders.
End users
- Site safety managers, supervisors, continuous improvement teams, and safety committees.

Pain points and challenges the product addresses

Manual observations and audits are intermittent and can miss repeated non-compliant behaviors.
Lagging indicators (injuries and incidents) are slow feedback mechanisms.
Organizations seeking leading indicators for coaching and targeted interventions may apply video analytics to those needs.

Qualification criteria: what makes a suitable fit

Adequate CCTV coverage of the highest-risk zones.
A mature safety governance process that can act on leading indicators (for example, coaching, engineering changes, signage, traffic plans).
Capacity to allocate time to camera scoping, rule setup, and tuning.

Real customer examples with company names when available

Marks & Spencer (referenced in vendor case study titles and in external coverage)
Procter & Gamble (listed in external coverage)
Hilton Foods (listed in external coverage)
3M (appears in a customer review context)

05The pitch

What the product does (core value proposition in objective terms)

- Protex AI is an AI video-analytics (computer vision) platform for workplace safety and operations that connects to existing fixed-position CCTV/IP cameras (via RTSP) and detects predefined unsafe behaviors and conditions (for example, PPE non-compliance, vehicle/pedestrian proximity, restricted-area breaches, ergonomics risks). It converts detections into event logs, dashboards, and workflows intended to help EHS and operations teams shift from reactive investigations toward proactive prevention activities. - The architecture described in vendor materials is hybrid: on-site edge processing for real-time analysis and privacy controls, combined with a cloud layer for analytics and reporting on event metadata and short clips.

Target market and market positioning

- Target buyers - Enterprise and upper mid-market EHS (Environment, Health & Safety) and operations teams in higher-risk industrial environments. - Target environments - Warehousing and logistics, industrial manufacturing, food and beverage manufacturing, and ports/maritime—environments where CCTV coverage is common and near-miss or behavior-based safety programs exist. - Positioning - The vendor frames the product as a tool to surface leading indicators from CCTV-derived behavioral and area/vehicle data, intended to complement incident reporting and manual audits.

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

Download full report →