Editor's Note: This blogpost was first published in May 2020, and has been revamped and updated for accuracy and comprehensiveness.

In 2018, industry research firm Gartner made an audacious prediction: that 85% of artificial intelligence (AI) projects won’t deliver. This means that amongst 20 artificial intelligence projects, three will succeed – leaving 17 falling short. According to Gartner, these projects are doomed due to bias in data, algorithms or the teams responsible for managing them. 

Unfortunately, the world is littered with failed AI and machine learning (ML) projects. It is easy for businesses to lose sight of the big picture in the midst of the nitty-gritty. Like any project, failed AI ventures have large cost implications alongside the high risk of reputational damage. 

In the same way that oil runs an engine, the fuel for AI and ML solutions is the data that feeds through it and the people that run it. Without the right quality or quantity of data, these technologies can’t do their jobs properly, or worse, become destructive due to inaccurate or harmful decision-making. Without the right quality of people and partners, the solution won’t scale and processes remain stagnant.

Overcoming barriers through partnerships ‍

Throughout the years, meldCX have been fortunate to head a number of successful AI projects, one example is our Concept SALi kiosk, done in collaboration with Australia Post. Short for Smart Automated Lodgement API, Concept SALi is a self-service kiosk that uses ML and computer vision technology to scan and detect parcels - verifying its size, weight, handwritten information, the sender’s identity, and shipping cost automatically. This would take the guesswork, manual entry and complexity out of shipping a parcel for customers and front of house staff in post offices. 

This project involved a wide variety of teams, both working locally and remotely across our Australian, Philippines and Taiwan offices - from industrial designers and hardware engineers, to software engineers, developers and UX/UI designers. Even though we’ve managed to achieve so much, this wasn’t a small feat. We owe our success largely to our close collaboration with strategic partners:

1. Ingram Micro - This year, we are thrilled to welcome Ingram Micro, one of the largest global distributors of technology products and services, as our channel partner for the Australia & New Zealand region. Linking up with Ingram Micro has been amazing, as they allow us to multiply our salesforce, working together to bring joint solutions to large enterprise customers. 

2. Cisco Meraki - As a leader in smart technology solutions, it has been our pleasure to partner with Cisco Meraki, as we offer Viana™, a vision analytics product by meldCX, bundled with Cisco Meraki MV cameras. What can we say…these products do work better together!

3. Microsoft - In the Microsoft Azure ecosystem, we use services in the IoT Hub such as Cognitive Services (Vision & Language) and IoT edge vision. Thanks to these technologies, our remote Dev Team was able to stand up an Azure machine learning solution in as little as two weeks. 

4. Intel - We worked together with Intel, utilising a combination of Intel technologies to power advance AI algorithms, including Intel Core i5 processors, Intel RealSense Depth Camera D435 for powerful image processing and depth perception, Intel Neural Compute Stick 2 for added performance boost, and the Intel Distribution of OpenVINO toolkit for AI optimisation and deployment. 

As both a Microsoft Partner and Intel Solutions Alliance Member, we’ve worked with them on a previous machine learning model, which served as building blocks for Concept SALi. Utilising Azure and Intel technologies on the SALi project lowered barriers of entry (e.g. for data migration), resulting in a shortened time to value.

5. Australia Post. One thing we love at meldCX is being able to drive greenfield and lighthouse projects together with customers. It was amazing to work with the talented team at Australia Post who had intimate knowledge of the problem statement, wanted to drive innovation for the business and change the paradigm using AI. This was a challenge we collectively accepted and worked as a team with great success.  

Building on each other’s strengths 

Building an effective ecosystem of partners becomes a fundamental part of a successful AI journey. Just as the saying by Hellen Keller goes, “Alone we can do so little; together we can do so much.” 

The relationship aspect is increasingly important for technology suppliers in the current phase of the IoT market. In conditions where everyone is learning as quickly as they can, there are moments where not everything will go as planned. Trust is the foundation of a successful partnership. Assembling a core team that are aware of each other’s capabilities, knowledge and resources become key. 

To achieve enduring traction and replicate successes, these relationships need to be continuously grown and built. By leveraging each other’s strengths, everyone wins.