Code Four

• Agency requires publicly available independent security certifications prior to procurement.
• Agency cannot accept cloud processing even in government-only cloud environments.
• Agency policy prohibits AI-assisted drafting of official narratives.

• Generated drafts are not fully autonomous and require human review and editing.

• No detailed public roadmap with delivery dates was found in the sources reviewed.

• No public benchmarks were found documenting maximum video hours per day, concurrency limits, or storage constraints.

• No publicly documented SOC 2 Type II or ISO certification found in the reviewed sources.
• Limited public detail on IAM controls (SSO/SCIM), logging, retention policy specifics, and key management.

• The vendor states broad integration intent, but terms indicate they may not perform automated writes back to CAD/RMS by default, which could limit end-to-end automation depending on agency needs.

• Video plus audio analysis capability is emphasized by the vendor and appears to be a differentiator relative to audio-only transcription tools.
• The vendor asserts support for multiple camera vendors (Axon, Motorola, Getac, and others are listed in vendor material), which can reduce hardware lock-in risk.
• The vendor references use of GovCloud hosting and asserts alignment with CJIS requirements in contract terms.

• Axon’s bundled ecosystem and market presence can reduce integration overhead for agencies already standardized on that vendor.
• Larger incumbents may have more extensive published collateral on responsible AI and security programs.

At the time of this research pass, Code Four does not have an easily verifiable presence on mainstream enterprise review platforms based on searches conducted. As a result, publicly available feedback is heavily weighted toward vendor-published testimonials, local press coverage, and community discussion.

Vendor-provided testimonials include statements attributed to police leadership highlighting rapid report catch-up during trials and high satisfaction among trial participants. One testimonial describes a positive service interaction involving an off-hours response from a founder; this is a vendor-curated testimonial and has not been independently verified in the sources reviewed.

A vendor case study for Pocatello Police Department states substantial reductions in report writing time and improved report comprehensiveness, while also acknowledging that AI-generated drafts require officer review and correction. Outcome metrics are vendor-reported and the case study does not disclose an independently audited methodology, sample size, or counterfactual controls.

Local news articles covering department pilots include comparative impressions between Code Four and alternative products in single-agency pilots. These reports provide useful anecdotal context but are not systematic evaluations.

Public discussion (e.g., local forums) includes skepticism about the extent of independent validation for accuracy and bias. Such discussions reflect public and practitioner concerns about reliance on vendor-published metrics without independent audits.

• To reduce report-writing time (vendor-reported metrics cited in case studies and testimonials)
• To analyze video plus audio rather than audio-only sources
• To avoid hardware vendor lock-in by supporting multiple camera systems
• To reduce cost versus some incumbent bundled solutions (vendor and local reporting suggest potential savings)

• Concerns about court defensibility and discovery/public-records implications
• Accuracy concerns for nuanced narratives
• Procurement risk associated with an early-stage vendor
• Integration limitations depending on RMS/CAD and evidence vendor

Claims: The case study reports substantial reductions in report-writing time and improved comprehensiveness, while noting that drafts require officer review.

Implementation detail: The case study emphasizes the importance of integrations and embedding the tool into existing workflows.

Critical evaluation: Reported metrics are vendor-provided with no independent measurement methodology disclosed. Error-rate statistics and legal/compliance outcomes are not reported.

What it adds: Confirms a real-world pilot where Code Four was compared side-by-side with an alternative solution.

Critical evaluation: Single-department pilot results are informative for local decision-making but do not necessarily generalize. Accuracy impressions from such pilots are useful anecdotal evidence but are not substitutes for systematic, independent audits.

• Agency size: Small-to-mid sized departments that experience report-writing backlogs and can pilot new tools quickly (often small municipal to mid-sized regional agencies).
• Environments with high body-worn camera usage and increasing public records demands.

• Decision makers: Chief/Sheriff, command staff, city/county IT leadership, procurement, legal counsel, records/public information officers.
• End users: Patrol officers, detectives, supervisors approving reports, and records clerks handling public records requests and redactions.

• Reducing time required for report writing
• Managing report backlogs
• Reviewing lengthy video evidence
• Redaction workload for public disclosure

• Agency willing to retain human-in-the-loop verification and maintain clear policy controls.
• Existing digital evidence cloud systems that can integrate with the vendor.
• An organizational and legal environment that permits AI-assisted drafting with transparent auditability.

• Primary buyers: US law enforcement agencies (police departments, possibly sheriffs’ offices)
• Positioning: The vendor presents the product as compatible with multiple camera vendors and as integrable with existing police workflows (RMS/CAD). The vendor markets the product’s video analysis capabilities as distinct from audio-only drafting tools.