01About Raven
What Raven is, and who it is for.
What is Raven?
Raven is Graylark's frontline visual intelligence platform. It analyzes images and individual video frames to return ranked investigative leads, including likely geographic locations, street-level candidate locations, vehicle identities, and image-authenticity assessments.
Who develops Raven?
Raven is developed by Graylark Technologies, the company that originally created GeoSpy.
Is Raven the same product as GeoSpy?
Raven is the evolution of GeoSpy. GeoSpy introduced Graylark's visual geolocation technology, while Raven expands that foundation into a broader visual intelligence platform for operational investigations. The full background is on GeoSpy is now Raven.
Who is Raven built for?
Raven is built primarily for law-enforcement, government, public-safety, intelligence, and investigative organizations that need to extract useful leads from low-context imagery.
What does "frontline visual intelligence" mean?
Frontline visual intelligence means turning imagery into information an investigator can act on. Raven is designed to help users move from an unknown image or video frame to ranked location, vehicle, or authenticity leads that can guide further investigation.
02Image geolocation
Working out where a frame was captured.
Can Raven determine where an image was taken?
Raven analyzes visual signals within an image to rank likely geographic locations. Depending on the image and available coverage, results may range from broader regional candidates to more precise street-level candidate locations. See Find Region and Find Street.
Does Raven require GPS or EXIF metadata?
No. Raven is designed to analyze the visible content of an image and does not require GPS coordinates, EXIF data, or other embedded metadata.
Does an image need to contain a recognizable landmark?
No. Raven can analyze less obvious visual signals such as terrain, vegetation, architecture, road design, materials, and environmental context when clear landmarks are absent.
How precise are Raven's location results?
Precision depends on the amount and quality of visual information available. Raven can return broader regional candidates or narrow an image to ranked street-level and precise candidate locations. Results should be reviewed alongside other investigative information.
Does Raven provide only one location result?
Raven can return ranked candidate locations so investigators can compare the most likely possibilities rather than relying on a single unsupported conclusion.
Can Raven analyze low-quality or partial images?
Raven is designed for low-context imagery and can often extract useful signals from partial or imperfect frames. Resolution, compression, obstructions, lighting, and the amount of visible environmental detail can affect the quality of the results.
03Video and other capabilities
Beyond location: vehicles, frames, and authenticity.
Can Raven analyze video?
Raven can analyze individual frames extracted from video. Each frame is treated as a still image and runs through the same capabilities.
Can Raven identify vehicles?
Yes. Raven's vehicle-identification capability analyzes partial interior and exterior visual signals to return ranked make, model, generation, and year candidates. See Identify Car.
Does Raven need a clear view of the entire vehicle?
Not necessarily. Raven is designed to work with partial, obstructed, interior, and exterior vehicle views, although clearer and more complete imagery can improve the usefulness of the returned candidates.
Can Raven detect AI-generated or manipulated images?
Raven can assess imagery for signs of AI generation and manipulation, including deepfakes and face swaps. The result should be treated as an analytical assessment and reviewed alongside the source, context, and other available evidence. See Verify Image.
Beta Image verification is currently offered as a beta capability.
What results does Raven return?
Depending on the capability used, Raven can return ranked geographic regions, street-level candidate locations, vehicle candidates, and image-authenticity assessments. Raven's outputs are designed to support investigation rather than replace investigator judgment.
See it run on your own imagery.
A demonstration is the fastest way to judge what Raven returns on the kind of frames your team actually works with.
04Accuracy and responsible use
How Raven's output should be read.
Are Raven's results guaranteed?
No analytical system can guarantee that every result will be correct. Raven returns ranked investigative leads based on the visual information available. Users should validate relevant results through additional evidence and established investigative procedures.
What affects the quality of a Raven result?
Relevant factors can include image resolution, compression, lighting, obstructions, field of view, visible environmental detail, geographic coverage, and whether the available visual signals are distinctive.
Are Raven's scores probabilities?
Raven may rank or score candidate results, but those values should not automatically be interpreted as calibrated probabilities unless the interface explicitly defines them that way.
Does Raven replace an investigator?
No. Raven helps investigators identify and prioritize leads. Final conclusions should incorporate human review, corroborating information, agency policy, and applicable legal requirements.
Can Raven's output be used as evidence?
Raven provides investigative leads and analytical outputs. How those outputs may be used depends on the surrounding evidence, agency procedures, applicable law, and the requirements of the relevant proceeding.
05Privacy and customer data
What happens to imagery you submit.
These answers summarize the published Privacy Policy and Terms of Service. Where the two differ, those documents govern.
Does Raven use customer imagery to train its models?
No. Customer imagery submitted to Raven is not used to train Raven's models. As stated in the Privacy Policy, Graylark does not use your images or metadata for AI training, further processing, or any other applications beyond delivering the services you request.
Does Raven permanently store uploaded imagery?
Raven is designed to process customer imagery without adding it to a permanent product training dataset. Uploaded images are retained only in line with the purpose of the search: the Privacy Policy sets out the current retention periods for quick searches and for images attached to a project or case file, along with how deletion works.
Who can access customer data?
Access to customer information is governed by the organization's account configuration and Graylark's applicable security and contractual controls. Contact Graylark for security and procurement documentation.
Where can an agency obtain Raven's security documentation?
Security, privacy, and procurement documentation can be requested directly from Graylark during an evaluation or purchasing process. Request a demonstration or email info@graylark.io.
06Deployment and integration
How teams get access to Raven.
How is Raven accessed?
Raven is used through a secure, browser-based workspace where investigators upload a frame and review the returned candidates. As set out in the Terms of Service, Raven may also be accessed through a mobile application or programmatically through Graylark's API.
Does Raven offer an API?
Programmatic access through Graylark's API is one of the access methods covered by the Terms of Service. Integration options depend on the organization's requirements, so contact Graylark to discuss current API and deployment availability.
Can Raven support enterprise deployments?
Raven can be evaluated for team and organization-wide use. Licensing and deployment structure depend on the intended users, workflow, and organizational requirements.
Is on-premises deployment available?
Deployment requirements are evaluated with the organization. Contact Graylark to discuss current hosting and deployment options.
07Evaluation, training, and purchasing
Starting an evaluation.
Can my organization request a Raven demo?
Yes. Qualified organizations can request a demonstration through the Raven website. The Graylark team will discuss the intended use case, users, technical requirements, and potential next steps.
Are Raven trials available?
Trials may be available to qualified organizations following an initial discussion. As noted in the Terms of Service, trial access is subject to prior review and approval, and its structure depends on the organization and intended evaluation.
Is training required?
Graylark may require onboarding or training before providing evaluation access so users understand Raven's capabilities, limitations, and responsible use.
How is Raven licensed?
Raven may be licensed for named users or broader organizational use, depending on the intended deployment. Contact Graylark for a structure appropriate to the organization.
How can an agency purchase Raven?
Start by requesting a demonstration or contacting Graylark. The team will review the use case, intended users, deployment requirements, procurement path, and appropriate licensing structure.
Can resellers or technology partners offer Raven?
Graylark works with selected partners. Organizations interested in reseller, integration, or strategic partnerships should contact Graylark directly.
