Darren Levine, Vice-President, Information Technology, Contrans Corp
A modern, thriving workplace is what great leaders aim to provide. A mindset of constant evolution will foster success for your organization, and the people that support it. Technology vendors often pitch solutions to common struggles, such as payment processing, and cash application. However, few vendors are solving the unique demands of your business. IT departments are in the best position ever to serve as strategic business partners that all organizations need to adapt and grow. Increasing pressure to maximize shareholder value, satisfy ever-evolving customer demands and process optimization can be solved with Artificial Intelligence (AI) and Machine Learning (ML).
Unique business challenges are typically found at the core of your operations where domain knowledge exists. Harnessing the power of AI can streamline, optimize, and in many cases, completely automate work that is repetitive, time-consuming, error prone, or a risk to your business. Your human capital can be freed to drive business growth and spend more time on strategic initiatives.
Getting Started
Implementing AI doesn’t need to be complex or time consuming. A vital component to successful implementations starts with a leadership team that fully supports and is engaged in your initiative. Successful projects are driven through participation. If stakeholders are not present and collaborative, find ways to pull them in, such as requesting input with meeting agendas and regularly asking probing questions. Roadmap the project in relatable and visually impactful ways that matter to your team. They need to easily see their challenges fitting inside your solution. Gain buy-in by generating benefit statements that are catered to each user’s perspective. Most importantly, recruit team members who perform the workload you plan to automate.
Change is difficult, uncomfortable, and the unknown is frequently scary for the people impacted by technology. They will naturally think what’s in this for me, or what will happen to my job? Clear communication helps counter those fears and reinforce that your project is meant to free people, allowing them to perform more value-add and higher-functioning work. Keep everyone engaged and reinforce benefits and opportunities. Ultimately, you are striving for a frictionless solution that will be entrenched into everyday workloads.
Finding Talent
It’s no secret that students and recent university graduates bring fresh perspectives and the latest tech knowledge to your workplace, making them valuable assets in your quest for automation. They think differently, are often more open to change, new ideas and new approaches.
A vital component to successful implementations starts with a leadership team that fully supports and is engaged in your initiative. 
Unencumbered by legacy systems and processes, students are more likely to question the status quo, suggesting new insights and creative ways of doing things. Their immersion into the latest technologies, trends and their natural inclination towards innovation and experimentation provides inherent advantages.
Foster their natural curiosity, focus on outcomes instead of activities, set guidelines rather than expectations. From personal experience, we’ve had tremendous success pulling in people from various education tracks and vocations to develop best in class solutions.
Adherence to your company’s standard operating procedures, such as documentation and change management is still required. These standards don’t need to be part of your initial focus. After providing time to explore and progress, introduce the team to the formalities of the business world. Lastly, find talent that is passionate and entrepreneurial to generate the momentum needed to create successful AI solutions.
Funding
Grant money is usually available through government programs and industry groups, which is how we initiated our first AI project. Finding the money and completing necessary paperwork may seem like a barrier, but students are more than capable of completing the requirements. Coop and intern programs often qualify for government subsidies too. Speak with your Payroll and HR departments to see what programs may be available.
Tools
Development tools don’t need to be costly. Choose a flexible and modern framework, such as:
• Microservices architecture
• Python backend
• JavaScript frontend
• ReactJS framework
Leveraging existing open-source packages, supported by tech innovators like Microsoft and Google, facilitates the development of AI solutions catered to the nuances of your industry. Some of the benefits of creating your own AI include removing monthly recurring fees from vendors and limiting license requirements. Perhaps most importantly, you retain control of your data. Instead of vendors having access to your proprietary information and profiting from it, you get to choose how it is used and who gets access to it.
Creating Scope and Setting Expectations
Striving for the perfect scope of work can pigeonhole your project. AI & ML is, unlike building an accounting interface, where everything must perfectly line up. In our first iteration, we clearly set goals to develop a solution that provided 85 percent accuracy and reported on areas for improvement. We were building a model to closely resemble real-life operations. However, we took the opportunity to ensure only critical processes, inputs and constraints were evaluated to minimize AI processing time.
Once we hit our first milestone, we realized that striving for perfection was a diminishing return. With an agile mindset, we pivoted to create an add on for easily managing and correcting exceptions before uploading final results into our core systems. Making the process iterative greatly shortened the time to value. We may attempt to automate to a 100 percent solution in the future, but user engagement skyrocketed when they realized the tool was interactive and they still retained control.
Measure Your Achievements
Building KPIs to monitor usage and success rates of your AI is critical. We built a planning tool for our first solution, measuring how frequently the tool was used and how many of the AI plans exactly matched what was uploaded into our core systems. These KPIs helped shape future iterations because we asked end users questions like why are you not running the AI more frequently and why did you alter the AI output?
Lessons Learned
AI and ML are revolutionizing the way we work, buy-in is key, funding is accessible, talent is available, and opportunities are near limitless, such as:
• Optimizing processes
• Improving productivity
• Increasing efficiencies
• Reducing waste
• Improving product quality
• Increasing asset utilization
• Increasing customer satisfaction
• Enhancing decision-making
Your journey to becoming a trusted business partner, proving your ability to adapt, continuously improve and thrive starts by developing creative and innovative AI solutions. I would like to give a big thanks to #ZakBrookshaw, the university grad who pioneered our first AI from concept to reality. We couldn’t have succeeded without his drive and passion!