# LANARS

> LANARS is a reviewed AI agency in Oslo, Norway, listed in the AIAgencies.eu register and last assessed 2026-09-09.

Canonical page: https://www.aiagencies.eu/agency/lanars
Generated: 2026-09-28 · Curated by Gabor Kiss · No paid placements — inclusion, score, and rank are never sold.

## Facts
- Location: Oslo, Norway
- Website: https://lanars.com
- Hourly rate: €45-85/hr (directory-listed)
- Team size: 10-49
- Founded: 2016
- Services: AI Development, AI Consulting, AI Automation
- Industries: Manufacturing, Healthcare, E-commerce
- LinkedIn: https://www.linkedin.com/company/lanars/

## About
LANARS is an Oslo hardware and software firm founded in 2016, pairing machine learning with embedded and edge engineering. It builds computer-vision systems that run on devices in the field, alongside IoT platforms and mobile applications.

## Documented clients
Documented engagements: SelectAI, Defigo, and Liquid Tool. All 3 carry a source URL on file.

## Curator assessment
Reviewed 2026-09-09 by Gabor Kiss against published evidence.

The case studies are better built than most, and the strongest one is not AI. Defigo documents an eight-year engineering partnership: Flutter apps, a Node.js and WebRTC backend at sub-second video latency, C++/Qt firmware on the control unit, 65,000 daily users, 3,000 installations and 55 million door openings.

The AI flagship is SelectAI, an in-cage vision unit for Norwegian salmon farms—three synchronized cameras under multispectral lighting, on-device pre-processing feeding cloud inference, more than 15,000 fish analyzed, classified against the regulator's own five-class lice taxonomy with no internal remapping. Live pilots today, with automated-counter approval targeted for end-2026.

The division of labor is published rather than blurred: LANARS wrote the firmware, the embedded vision stack, the annotation platform, the cloud path and the over-the-air pipeline, while model architecture, training and validation sat with a partner, AI Experts. That is the thing to probe.

The weaker fit is a buyer who needs the models themselves owned in-house by the supplier.

### Key strengths
- The SelectAI write-up separates what each party built, naming the partner that owned model architecture, training and validation
- Defigo documents an eight-year engineering partnership with running numbers: 65,000 daily users, 3,000 installations and sub-second video latency
- Edge constraints are treated as the design problem—three synchronized cameras, multispectral lighting and on-device pre-processing feeding cloud inference

### Good to know
- Model work on the flagship AI project belonged to a partner—ask who would train and validate models on your engagement
- SelectAI is described at live-pilot stage with regulator approval targeted for end-2026; ask what runs unattended today
- 36 people on LinkedIn against offices in Oslo, Edinburgh and Vilnius—confirm which site staffs your project

---
Source: AIAgencies.eu — the curated register of European AI agencies. Rate provenance: "agency-confirmed" = disclosed via submission or claimed profile; "directory-listed" = Clutch or two agreeing B2B directories (lone unverified sources are never written). Full method: https://www.aiagencies.eu/methodology · Rate index: https://www.aiagencies.eu/rates