Alfred & Fovea — AI Agent & Computer Vision Products | LunaVIs

Alfred & Fovea.

Two proprietary AI products. Alfred handles business automation — Fovea handles computer vision intelligence. Both deploy as standalone platforms, not bespoke builds.

Built on open foundations: LangGraph, TensorRT, WhatsApp API, multi-tenant cloud architecture. Currently in Early Access — onboarding select pilot clients across MENA.

Early Access StageEarly access pilots open · MENA & beyond
AI Agent Platform

Alfred

Customizable AI Agent — deployed as a platform, not a project

Alfred is a skill-based AI agent platform. Each Alfred deployment is configured with the skills your business needs — lead qualification, appointment booking, CRM sync, multi-step workflow orchestration. One agent, fully autonomous.

WhatsApp Digital Receptionist
Qualifies leads, books appointments, and answers FAQs around the clock.
Skill-Based Agent Architecture
Alfred's skills are modular — each skill handles a specific business function via LangGraph.
Context-Aware AI Memory
Builds a persistent knowledge layer over every client interaction.
Automated Reporting & CRM Sync
Zero manual data entry — everything logged and synced automatically.

Early Access · Pilot Program

Core Alfred agent framework and multi-skill pipeline infrastructure built. Currently onboarding pilot clients in Pakistan and Egypt, with plans to scale globally.

Request Alfred Access
Alfred · Agent DashboardAgent Active

WhatsApp · Incoming Lead

Hi, I'd like to book a viewing for tomorrow at 3pm

10:42 AM

LunaVIs AI Agent

Hi! I've checked availability — tomorrow 3pm works perfectly. I've sent a confirmation to your number. Is there anything else you need?

10:42 AM · 1.3s

Perfect, thanks!

10:43 AM

24/7
Response
< 2s
Reply Time
100%
Auto-booked
Fovea · CV Dashboard · Zone BCV Pipeline Active
LIVE
14:32:01Unauthorized entry detected — Zone B entrance
14:31:44Person dwell time: 4m 12s — Aisle 3
14:31:20Re-ID match: Person #07 re-entered via Gate 2
4
Cameras
~30ms
Latency
Real-time
Alerts
Computer Vision AI

Fovea

Computer Vision Intelligence Platform — edge & cloud

Fovea is a real-time computer vision platform that monitors physical spaces, identifies anomalies, tracks movement patterns, and triggers automated responses — without a human watching a screen. TensorRT-powered for low-latency edge deployment.

Multi-Camera Tracking & Re-ID
Track individuals across multiple camera feeds with persistent re-identification.
Event-Triggered Automation
Unauthorized entry → WhatsApp alert → incident log, instantly.
Retail & Space Analytics
Heatmaps, dwell-time stats, and zone analytics for smarter layouts.
TensorRT + DeepStream Powered
GPU-accelerated inference pipeline for ~30ms real-time latency.

Early Access · Pilot Program

Core Fovea detection and tracking pipeline complete. Multi-camera re-ID and real-time alerting in active development. Retail and security pilot deployments targeted for Q2 2026.

Request Fovea Access

Proprietary stack. No contractor toolkit.

Alfred and Fovea are built on a purpose-built infrastructure — open foundations, proprietary orchestration, multi-tenant by design.

LangGraph
Alfred · Orchestration
FastAPI
Backend API
TensorRT
Fovea · CV Inference
DeepStream
Fovea · Video Pipeline
InsightFace
Fovea · Re-ID
PostgreSQL
Database
Qdrant
Alfred · Vector Store
Docker
Containers
GCP / GKE
Cloud Deploy
WhatsApp API
Alfred · Messaging
Ollama
Alfred · Local LLM
Cloudflare
Edge / CDN

Interested in a pilot?

We're onboarding select early-access clients. MENA-first, scaling globally.

Request Early Access
LunaVIs
Assistant
Online