Company
Building the capture layer for AI-ready automotive operations.
Halo Eye is building Halo Compass, browser-based capture infrastructure for automotive workflows where visual evidence is critical. The platform guides any operator through standardized vehicle capture at the source, using computer vision, real-time quality control, and workflow-driven prompts. In ~90 seconds, Halo turns one guided session into listing-ready imagery, close-range evidence, and structured visual outputs — without apps, hardware, or dedicated photo bays.
Operational proof
The thesis
The bottleneck is not always the model. It is the capture.
Automotive workflows increasingly depend on images, metadata, and machine-readable visual context. But most vehicle imagery is still created through inconsistent operator behavior, tight lot conditions, missed views, disconnected handoffs, and fragmented workflows.
Halo exists to make capture correct, repeatable, and useful before visual inputs enter downstream systems.
Inconsistent capture weakens every workflow built on top of it.
What we are building
A capture orchestration layer for vehicle workflows.
Halo Compass treats capture as a source event. It guides the operator, validates readiness before acceptance, and turns one session into reusable operational infrastructure.
Presentation imagery
Standardized views for merchandising, publishing, and customer-facing workflows.
Evidence imagery
Close-range visual proof that preserves condition context before teams and systems diverge.
Operational metadata
Readiness, coverage, session, and workflow context for downstream systems.
Founder-market fit
Built from inside automotive operations.
Halo was shaped by years of watching fragmented vehicle imagery weaken operational decisions.
Halo was founded by Stephen Southin, a lifelong automotive operator and product builder with 34 years in the industry.
Stephen’s previous work included building and commercializing AI-powered vehicle inspection infrastructure across high-scale fleets and some of the world’s largest enterprise clients.
Product philosophy
Better downstream intelligence starts upstream.
Halo does not begin with dashboards, reports, or after-the-fact image correction. It begins at the moment visual truth is created.
The product philosophy is simple: improve the input before asking automation to reason over it.
01
Capture first.
AI workflows depend on better visual inputs before automation begins.
02
Workflow-native.
Designed around how vehicles move through stores, fleets, service, recon, marketplaces, and handoff environments.
03
Browser-based by design.
No app-store dependency, no dedicated hardware, and no GPU requirement for deployment.
04
Structured for integration.
Every guided session can become an event that other systems can use.
Team & ecosystem
Built at the intersection of automotive operations, product systems, and applied machine learning.
Halo combines operator context with product design, computer vision, and platform infrastructure.
Automotive operations
Dealership, fleet, inspection, merchandising, and vehicle workflow reality.
Product & workflow design
Browser-based systems built for repeatable operator behavior.
Applied machine learning
Real-time guidance, readiness validation, and visual output structure.
Platform infrastructure
APIs, workflow context, metadata, and event-driven architecture.
Applied AI partners
Built in Toronto with support from a strong innovation ecosystem.
Built in Toronto with support from applied innovation and machine learning partners.
Long-term vision
Building infrastructure for AI-ready automotive workflows.
Halo is designed as infrastructure that can support AI-ready workflows across the vehicle lifecycle.
When successful, Halo becomes the capture layer that helps operational systems, automations, integrations, and AI decision environments reason from better visual inputs.
Operational infrastructure
Make capture correct before the workflow begins.
AI workflows only become reliable when the visual inputs become operationally trustworthy.