What we do

Object Identifcation
Fruit, animals, people, cars, products.. whatever it is you need to identify in an image or video. We love to work across a range of industries and problems. We are data scientists after all.
Object Classification
Group objects based on certain features like type, size, location. Think colour, damage, demographics, time of day. This is where the value of the insights become super powerful.
See Behaviours
With sufficient training, a machine can identify and predict patterns of behaviour. This is where is gets really exciting!
Empowering Predictions
Data science is all about predictions and we love them! Think of event you would like to predict and lets talk!
Business Intelligence
Organisations perform better with more informed decisions from data. Its likely you have a lot more useful data than you realise. Lets unlock it together.
Data Challenges
Its likely you are unable to make the most of your data because its either unstructured, messy or there is too much of it. We have the tools to overcome these challenges.


Leading machine learning in ANZ

Our Experience

We have been fortunate to work across a wide range of projects for enterprise, SME and government. This puts Intela in a unique position to confidently engage with opportunities and challenges across any sector or domain.

Our Team

We have an awesome team of data scientists, with broad machine learning experience and capability that covers research, engineering and deployment. We also partner with awesome local companies who are experts outside of our core focus.

Examples Of Our Work

Client: Transpower

Object Identification & Classification

Intela AI were engaged by Transpower, New Zealand’s electricity transmission system owner and operator, to help explore the potential applications of Artificial Intelligence to transmission line inspection work.

Intela AI’s exciting new machine vision capabilities analysed conductor videos and images for damage; then rapidly identify, classify, and report the identified defects.


This capability is predicted to significantly cut down the time and cost of transmission line inspection, while increasing accuracy of reporting. As part of their innovation strategy, this aligns with Transpower’s objectives to maximise the use of existing assets while minimising community and environmental impacts, in line with their commitment to ensuring that the National Grid meets the needs of all electricity consumers now and into the future.


The machine vision solution performed significantly better than human review in terms of identifying damaged cables and in the time taken to review images.

Internal Research

The Problem

Fruit growing and processing is a huge industry in New Zealand. Intela were made aware by a leading sector player that they only samples 5% of their tree’s for yield estimations because of the time it took for human counting. Given the significance of accurate harvest forecasts for pricing, supply chain and resource allocation; being able to sample 100% of the crop would add significant confidence for the business.

The Solution
  • Video fly-by of all tree’s using a camera mounted on a vehicle (tractor or ATV) or drone.
  • Machine learning model to process the footage and accurately count the fruit on each tree.

The final solution will enable the client to get a much more accurate picture of their 1million tree operation. With frequent fly-by’s, data could be collected to predict harvesting times and provide insights useful to the sales team and distribution customers.

The business intelligence potential could help transform the operations of the business.


We have so far been able to accurately identify and count:

  • Apples
  • Avocado
  • Strawberry
  • Mango
Fruit identification

Client: Government

Object Identification & Classification

Intela AI were engaged by a central government agency to help explore the potential of Artificial Intelligence to support its industry monitoring efforts.

  • Intela first researched and reported on the possible solutions, their viability and business requirements.
  • Next Intela built an image labelling platform for domain experts to easily support the training of machine vision models.
  • Finally Intela deployed the machine learning models that can rapidly analyze video footage to identify, count and classify the animals.

The final solution will enable the monitoring program to be scaled to analyze the estimated  1million hours of video footage that will be collected each year.

Without using AI, the agency would be faced with a huge overhead of people and systems to effectively monitor. Having machine review the video also allowed the agency to manage certain legal implications of having human reviewers.


The machine vision solution performed significantly better than human review in terms of the time taken to review footage.

Internal Research

The Problem

City planners, space designers, retailers and building owners all want to know how people interact with their environment. There is an abundance of video footage available from existing CCTV style cameras, but most do not have the capability to extract valuable analytic data.

New Zealand has uniquely strong anti-surveillance legislation, so the aim of the research is not to identify individuals, but to see patterns of behaviour and extract metadata. The strong privacy culture also drives the need to reduce the need for humans to review video; removing the possibility of bias or abuse entirely.

The Solution

Intela are experts in machine vision, research was initiated to demonstrate what ‘low hanging fruit’ could be picked off raw video feeds.

We rapidly worked up capability to report:

  • People counting
  • People direction
  • People movement tracking
  • Dwell times
  • Anonymisation filters

Rich intelligence can be acquired from existing video hardware, without having to compromise privacy.

Ready to realise the value of machine learning? Request More Information

Recent News

Intela is now Arcanum.
Jump over to the website to learn more about how we are helping make humans great.