AI-Powered
Surveillance
The right platforms for law enforcement, border security, critical infrastructure operators and everyday citizens. Together we make the world a safer place and hold perpetrators accountable. NesherAI is finally live! GeoLink and ShadowLink are launching in May 2026 with superior capabilities.
Perfected.
Search a billion scale database using a single photo — even when the subject face is highly masked, at an extreme angle, or captured in low quality. Build watchlists, run hundreds of face comparisons from a couple of images, track identities through extended video footage within milliseconds, extract and search identities from video files, monitor real-time face identifications through an integrated NVR hardware pipeline.
Determine the precise location of outdoor and indoor views from both images and video frames without any need for metadata. GeoLink analyzes visual input signals and returns high-confidence and high-precision coordinates from 190+ countries.
Upload millions of case files and watch ShadowLink extract every entity, location, face identity, and piece of evidence with deep visual and linguistic comprehension. People, organizations, devices, financial accounts, vehicles, phone numbers, email addresses, and crypto wallets — all identified, linked, synthesized into evidence trail and visualized in an interactive relationship graph. Reduces weeks and months of investigative work into minutes.
The single authenticated portal for all Stellar products. NesherAI, GeoLink, ShadowLink, NVR hardware monitoring, cloud storage, API keys, billing, and team management — centralized with role-based access control and a full audit trail.
Clear pricing for every team size.
NesherAI standard plans are self-serve with identity verification. Professional plans include hardware, large-scale deployments, and on-premises — contact sales for a quote. GeoLink and ShadowLink pricing available at launch in May 2026.
Basic $300/mo | Most Popular Pro $600/mo | Max $900/mo | |
|---|---|---|---|
| Searches | |||
| Daily Face Searches | 10 | 20 | 35 |
| Daily Video Searches | 0 | 0 | 5 |
| Collections | |||
| Search Collections | 1 (500 faces) | 1 (2,000 faces) | 3 (10,000 faces each) |
| User Collections | 1 (500 faces) | 1 (2,000 faces) | 3 (10,000 faces each) |
| Storage | |||
| Included Cloud Storage | 50 GB | 200 GB | 1 TB |
| Features | |||
| Master Database Search | |||
| Private Collection Search | |||
| Search Collection Access | |||
Built for the mission
Stellar Technologies builds intelligence platforms for operators who work in conditions that defeat commercial tools — degraded imagery, adversarial subjects, incomplete evidence, and no tolerance for false results.
NesherAI was built because existing facial recognition failed the moment conditions got hard. Masks, extreme angles, low-resolution CCTV, partial occlusion — most systems return nothing useful. NesherAI returns ranked matches with similarity scores from the most challenging source imagery. The platform now covers face search, face comparison, private collection management, video processing, and real-time NVR surveillance through a complete hardware-to-cloud pipeline.
GeoLink and ShadowLink launch in May 2026. GeoLink identifies where any outdoor photograph was taken using visual analysis alone. ShadowLink automates investigation workflows — extracting entities and mapping relationships from raw case files at a scale that takes an analyst team days to replicate manually.
Press Coverage
In March 2025, NesherAI was reported by AP News and Times of Israel in coverage documenting its deployment in real-world operational conditions — crowded outdoor environments, varied lighting, and partial occlusion.
Times of Israel reported on the use of NesherAI to identify individuals at campus protest events, documenting the platform's capability to match faces from low-quality crowd imagery in outdoor conditions.
AP News reported on the deployment of facial recognition tools including NesherAI in the identification of individuals at campus protest events, covering the system's use in real-world operational conditions.