Limbus Tracking Demo
Shows temporal limbus tracking in motion, highlighting stable iris boundary detection and real-time ocular geometry analysis.
VectorCall AI LLC designs, builds, and deploys AI solutions — from real-time clinical computer vision to blockchain security layers, spatial augmented reality, and LLM-powered systems.
VectorCall AI LLC is an applied artificial intelligence consulting and engineering firm helping businesses transform data into actionable intelligence. We specialize in designing, building, and deploying AI solutions that solve real-world problems across industries.
Our team brings deep expertise in computer vision, machine learning, blockchain AI security, and spatial augmented reality. We support organizations by developing custom models, automating workflows, and integrating AI into existing systems — moving projects from concept to production with measurable outcomes.
A real-time ocular geometry engine for clinical eye tracking. Uses radial gradient search and RANSAC ellipse fitting on the iris-sclera boundary with temporal gating — no training data or data collection required. Includes five ophthalmic screening tools: strabismus detection, dry eye / blink analysis, keratoconus risk scoring, micro-saccade / concussion screening, and surgical cyclotorsion alignment for LASIK and toric IOL procedures.
A private-beta AI security layer for stablecoin payments and agent-led commerce. Evaluates transfer risk, invoice fraud, deepfake-aware KYC signals, and agent-initiated spending before funds move — with plain-English decision explanations. Features a live playground, REST API, dashboard, and adversarial red-team audit surface. Currently integrated with Polygon testnet settlement flow.
A camera-projector spatial AR system that registers LiDAR-captured 3D point cloud scans (Scaniverse) to physical surfaces using homography-based calibration and depth-dot alignment. Projects virtual overlays precisely onto real-world geometry — enabling interactive augmented reality experiences without wearables, using only an off-the-shelf camera and projector.
An AI-powered chess game tracker with a camera mounted above a physical board. YOLO-based piece detection automatically recognizes every move in real time, announces it with voice narration (including move timing), visualizes the live board state in a web interface via WebSocket, and stores full game history in a SQLite database — zero manual input required.
A spatial AR piano coaching system that listens to MIDI input from a digital piano, tracks hand position over the keyboard, projects guided note prompts onto the keys, and delivers real-time voice and LLM-assisted feedback. The platform combines projection mapping, hand tracking, fingering analysis, song libraries, and a local dashboard to keep a learner focused on the instrument instead of a separate screen.
A browser-based clinical annotation tool for marking 21 anatomical keypoints on cleft lip images. Features include zoom, pan, undo, per-keypoint selection with visual confirmation, adjustable marker size, and CSV export of pixel-accurate coordinates. Built for craniofacial research teams in plastic and reconstructive surgery to generate training data for AI-driven surgical planning and outcome assessment.
A local video annotation tool for frame-by-frame labeling of velopharyngeal insufficiency (VPI) and velopharyngeal patterns (VPP) during nasopharyngoscopy evaluations. Clinicians mark start/end timestamps of positive or negative events directly on the video timeline, with color-coded visual overlays on the scrubber. Annotations export as structured metadata for AI model training — supporting cleft palate speech surgery research and VPI classification systems.
A cross-platform dating application built with Flutter and Firebase. Features Tinder-style card swiping, mutual-like matching that auto-creates chat rooms, real-time messaging with read receipts, and a 13-field profile system covering interests, relationship goals, lifestyle, and demographics. Includes safety guardrails (block/report with moderation pipeline), smart ad monetization with daily caps, and a 200-user fake-data seeder for development. Runs on Android, iOS, Web, Windows, macOS, and Linux from a single Dart codebase.
Shows temporal limbus tracking in motion, highlighting stable iris boundary detection and real-time ocular geometry analysis.
Represents the live data flow behind our model pipelines and decision engines.
Reflects depth mapping, projection alignment, and real-world AR registration.
Highlights continuous risk scoring, fraud analysis, and transaction safeguarding.
End-to-end model design, training, optimization, and deployment for detection, classification, segmentation, and prediction tasks tailored to your domain.
Real-time image and video analysis pipelines — from camera calibration and object tracking to clinical-grade ophthalmic screening and surgical guidance tools.
Intelligent security layers for Web3 workflows, stablecoin fraud detection, KYC signal analysis, and agent-initiated transaction safety at the protocol level.
Production-grade LLM-powered applications with robust prompt engineering, retrieval-augmented generation, adversarial audits, and safety guardrails.
Whether you have a project in mind, need a technical partner, or want to explore what AI can do for your business — we'd love to hear from you.