Your security is
our no. 1 priority.

Absolutely. As a service partner to organizations across the defense and pharmaceutical sectors, data security is a priority at Zigzag AI. We follow a multi-layered security approach across the network, data, and application layers. Our safeguards include AES-256 encryption at rest, HTTPS/TLS encryption in transit, role-based access control (RBAC), secure authentication, and client data isolation. Access is restricted to authorized users, with activity logging for traceability. Data retention, backup, download, and deletion are managed according to agreed project requirements and policies.

Zigzag AI follows a structured approach to managing cybersecurity risks, protecting client data, and maintaining operational continuity. Our practices include regular Vulnerability Assessment and Penetration Testing (VAPT), data protection measures aligned with applicable DPDP requirements, and internal disaster management and incident response training. We continuously work to identify potential vulnerabilities, strengthen our security controls, and improve our readiness to respond to security incidents and operational disruptions.

Yes. Zigzag AI is ISO/IEC 42001:2023 certified, demonstrating our commitment to structured AI management, responsible AI practices, and continuous improvement in how we develop and deploy AI solutions.

No. Zigzag AI follows client data isolation practices to help ensure that each client's data remains separate and accessible only to authorized users. Additionally, we maintain Non-Disclosure Agreements (NDAs) with clients to protect confidential business information and ensure it is handled responsibly.

Zigzag AI offers flexible data storage options based on your organization's requirements. In regular practice, data are stored in a Zigzag AI-managed secure Amazon S3 and Amazon RDS cloud environment. Data access is restricted to authorised personnel and services, with client data isolation controls designed to prevent access by other clients. For organizations requiring greater control over their data, deployment on their own cloud infrastructure or on-premises servers is also available at an additional cost, depending on the infrastructure and deployment requirements.

By default, Zigzag AI retains data for 45 days. However, the retention period can be increased or decreased based on your organization's requirements and agreed project configuration, giving you flexibility over how long your operational data is stored.

No. Client data is used to deliver the agreed services, generate operational insights, and improve performance for your organization's specific use cases. Your business data and video footage are not used to train AI models for other clients.

In most cases, NO. One of Zigzag AI's key advantages is that it works with your existing CCTV infrastructure, helping you avoid unnecessary investment in new cameras or hardware. The solution integrates with existing IP cameras and DVR/NVR systems through RTSP video streams.

Zigzag AI can work with existing IP cameras across different resolutions, so you generally don't need to upgrade your cameras solely for AI deployment. For optimal performance, we recommend the following specifications: Resolution: 4 MP/5 MP/8 MP Camera Type: IP cameras are recommended. Camera Design: Bullet cameras are preferred. Dome and fisheye cameras are not recommended.

Zigzag is camera-brand independent and works with standards-based IP cameras and DVR/NVR systems that provide accessible RTSP streams. CP Plus is recommended as the preferred brand.

Bandwidth requirements depend on the camera's encoded bitrate, resolution, and video settings. For H.264 video, a typical planning range is approximately 2–8 Mbps per active camera while scheduled or relevant video clips are being captured. Actual requirements may vary based on your camera configuration and deployment scale.

Yes. Zigzag AI supports all three deployment options: Cloud, On-Premises, and Hybrid. Cloud deployment is generally recommended for faster onboarding, scalable AI processing, centralized monitoring, automatic updates, and access to the complete analytics, alerts, dashboards, and reporting features without managing dedicated infrastructure. On-Premises and Hybrid deployments are also available for organizations with specific infrastructure or data control requirements. These options typically involve higher initial investment and maintenance, making them more suitable for larger-scale deployments where the operational benefits justify the cost.

Hardware specifications requirements depend on the deployment scale, number of cameras, video resolution, and analytics use cases. Zigzag AI team will assess your requirements and recommend the appropriate configuration. For simpler deployment, we recommend Cloud hosting, which eliminates the need to procure and maintain dedicated GPU/CPU infrastructure.

Zigzag AI can monitor a wide range of operational activities across manufacturing, hospitality, retail, healthcare, pharmaceuticals, logistics, and other industries using existing CCTV cameras. Capabilities include safety and SOP compliance, workforce productivity, machine utilization, process-flow monitoring, Quality inspection, Automatic Number Plate Recognition (ANPR), housekeeping, customer and staff interactions, vehicle movement, security incidents, and more. Customization is at the core of Zigzag AI. You can define what you want to monitor in plain language, and our team configures the relevant AI use cases, alerts, dashboards, and reports around your specific operational requirements.

Zigzag AI typically targets an initial detection accuracy of approximately 70–75% during the initial implementation phase. Following the successful completion of the POC, our team continuously fine-tunes and optimizes the AI models through phased implementation and feedback. Our goal is to achieve over 97% detection accuracy, wherever technically feasible, depending on the use case, camera positioning, lighting, video quality, and deployment conditions.

Yes. Zigzag supports near-real-time analysis with immediate output generation.

Multiple AI models can run concurrently on a single camera stream. The platform uses a scalable pipeline architecture, and concurrency is sized according to complexity, requirements, sampled FPS, and the required response time.

Yes. A single camera can monitor multiple activities, use cases and defined zones simultaneously, depending on the camera's field of view, video quality, and the selected AI use cases. Zigzag AI can be configured to monitor different operational activities within the same camera view, with customized alerts, dashboards, and reports based on your requirements.

Yes. Zigzag AI offers a Proof of Concept (POC) so you can evaluate how well it works for your specific operational requirements before making a commitment. This allows you to validate the impact of AI generated detections, alerts, dashboards, and reports using your existing CCTV infrastructure and make an informed decision based on the results.

Zigzag AI typically goes live within 1–2 business days after receiving the required operational requirements and completing all prerequisite connections. Our team handles the setup and configuration, so your team does not need to be involved in the implementation process.

Zigzag AI offers some of the lowest pricing in the industry, with plans starting at less than 1K per camera per month. Get AI-powered operational insights, real-time alerts, and customised dashboards without a heavy investment and new infrastructure. For pricing tailored to your requirements, contact our team at team@zig-zag.ai

Zigzag AI can notify the relevant teams when a configured incident or operational deviation is detected. Notifications can be delivered through supported channels such as WhatsApp, email, and voice-based announcements using the Zigzag Cue alert system, helping teams respond promptly to operational issues.

Yes. Alerts can be configured according to your operational requirements, including the type of incident, monitored area, and applicable rules or thresholds. This helps teams receive relevant notifications without unnecessary alerts for every detected event.

Zigzag AI provides customized reports and dashboards based on your operational requirements and selected AI use cases. Reports can be generated in Excel, PDF, or web-based formats, while dashboards can be designed in 2D or 3D, depending on your requirements and the nature of the data being visualised. The format, metrics, and presentation can be tailored to suit your team's monitoring and reporting needs.

Yes. Customization is at the core of Zigzag AI. Reports and dashboards can be tailored to your processes, monitored activities, reporting metrics, locations, and management requirements. You can define what you want to monitor, and our team configures the relevant analytics and reporting accordingly.

Yes. Reports can be configured for scheduled delivery at agreed intervals, such as daily, weekly, or monthly, depending on your requirements and deployment configuration.

Yes. Depending on the report and deployment configuration, operational data can be made available in supported export formats for further analysis, record-keeping, and internal reporting.

Zigzag AI can be configured to work with existing business workflows and systems, depending on the integration requirements and technical compatibility.