Putting AI to work


Artificial Intelligence (AI) Technologies offers vast opportunities for businesses that are willing to embrace this revolutionary change, to name a few opportunities: cost efficiency, operational effectiveness, agility, delighting customers and ultimately richer customer experiences. Technology tends to provide evolutionary (incremental) changes in the short term and revolutionary (transformation) in the long term. The benefit with AI is that it brings transformative changes for organisations that are able to use it to their advantage. In most cases, these technological changes are visible internally and externally, i.e. improved company bottom line and the latter, convenient, quality products and services for consumers. Unfortunately this technological transformation can have positive and negative outcomes though we can be better prepared for it if we understand what is possible or not and if we plan in advance. For organizations this means having a clear vision and coherent digital transformation strategy, i.e. fulfilling customer value proposition using new business models, adopting agile and innovative culture across the organization, cultivating talent management and using technology to enable competitive advantage.

Organizations application of AI technologies is still in its infancy stage, as in most cases many organizations may have limited budget for this. Though budget is not the only challenge, AI talent is still scarce given that it requires advance specialization in Big data and analytics, data science and software development. Some organizations are still experimenting with AI, these may have sufficient resources to do so. It is important to note that not all companies are able to experiment with AI, especially small to medium businesses, unless these businesses are using AI to differentiate themselves (specialization) and as a competitive advantage in the market. Based on my experience I see a similar pattern with the adoption of most emerging technologies, i.e. 5G, blockchain etc.

Where to start to apply AI for enterprises?

The first step is to imagine or re-imagine the business landscape, internal and external environment scanning, i.e. organizations need to ask themselves the following questions:

  • Is AI part of our business strategy?

  • What technologies can we use, existing and emerging?

  • Which systems and processes are we improving?

  • What product and services will be enhanced? and

  • What is the value proposition to customers?

AI initiative should be governed by AI strategy which in turn must be aligned with the business objectives and overall company strategy.

As a senior IT capacity planner, we used to collect performance data for IT infrastructure systems, i.e. servers, storage arrays etc. using a capacity planning tool. This data will be housed in a Database or Datamart. We analyzed the data to identify system performance issues and to forecast future demand using variables like, CPU, disk, network utilization, number of users and volumes, to name a few. In doing so, we created models and used different analytics algorithms embedded on the tool. The more data we collected the better the patterns and insights and vice versa. So how does this relate to AI? I see a common thread with how AI works. AI is mostly data and analytics, whether it's in form of numbers, text or pictures. AI should be an extension of already existing data warehouse and analytics. Most organizations collects vast amount of data on a daily basis. These organizations have an advantage to start with an AI initiative. Since working with data is a sensitive topic, especially customer data. Organizations needs to ensure they adhere to ethical standards, governance and in-country regulations.

The second step is to invent or re-invent customer value proposition, aligning with the 'what' (objectives) and 'why' (vision and mission) for the organization - defining what good looks like. This may be integrating AI in products and services to improve overall customer experience or integrating AI in internal processes and systems for efficiency and operational effectiveness. The reasons will vary for different organizations. Subsequently the 'how' to achieve vision and business objectives should be defined (AI strategy). This may include improving customer experience, creating smart products, adopting new digital technologies such as cloud computing, big data and analytics, AI technologies, new ways of working and addressing skills gaps. In some instances partnering with other businesses to complement internal skills or acquiring AI startups etc. During this step it is important to define clear key performance indicators (KPIs) and critical success factors (CSFs) to measure success.

The third step is the most important, ignite AI strategy. Most research agree that most initiatives or strategies fail in implementation for various reasons. Most organizations are adopting agile methodology to deliver a minimum viable product (MVP), fail fast and learn quickly whilst delivering incremental value to customers.

Adopting lean startup methodology by Eric Reis from design to market products faster can maximize value not only for startups but also for enterprises. Lean startup methodology uses Build-Measure-Learn approach to turn ideas into products, measure how customers respond, and then learn whether to pivot or persevere. These are just a few proven methods that organizations can consider to realize their strategic objectives. Our iterative strategic thinking process uses different models, frameworks, concepts and methodologies based on the business opportunity or challenge.

The fourth step is continuous improvement and innovation, measure and re-imagine. This can be achieved by measuring defined KPIs and CSFs. For example, are we achieving business objectives in line with the dynamic business environment and ever-changing customer needs? If the strategy is not delivering the intended results, leadership must be willing to pivot and re-imagine the value proposition.

Five application (use cases) of AI, Machine learning and Deep learning for businesses

The aim of these use cases is to raise awareness of digital innovations happening in AI, machine learning, deep learning and how businesses can take advantage of these advances. To realize business value, these technologies must have specific use cases that deliver measurable business outcomes. AI strategies should be grounded by business objectives. Common AI Technologies:

  • Machine learning is best suitable for structured numeric data (statistics);

  • Natural language processing, for text and speech recognition;

  • Deep learning for images and

  • Robotic Process Automation (RPA) to automate manual and repetitive tasks.

1. Continuous Improvement - Optimize and automate business processes and applications

RPA is often referred to as digital labour or synthentic worker since it mimic a human. It is essentially computer programs known as robots though many would say this is entry level AI. This technology is improving as with the adoption into mainstream. AI/Machine learning can be used to automate business process and applications. The benefits of AI in this space are obvious, to name a few:

  • Lower operational cost,

  • Reduced error rates (human error) through automation,

  • Automating manual and repetitive tasks,

  • Increased employee productivity,

  • Increased agility, i.e. speed to market and

  • Ultimately delighting customers.

Amazon Go convenience stores using fully automated checkout is a good example for AI application. According to Amazon this is the world's most advanced shopping technology. No lines, no checkout - just grab and go.

2. Virtual assistant and conversational user experience

Voice operated assistants are beginning to replace a variety of tasks. For example, my two year daughter likes to use the voice feature powered by AI Google assistant on my smart phone, i.e. she will say "play Wendy and Uncle John". Her voice command will then direct Google assistant to open Youtube and play the video. Wow, isn't that amazing! This natural language processing application is good at answering questions. Virtual assistants are a game changer. Companies are looking at how to integrate this technology on their products to improve user experience.

3. Facial recognition

A facial recognition system is a technology capable of identifying or verifying a person from a digital image or a video frame. It is a Biometric Artificial Intelligence based application that can uniquely identify a person by analyzing patterns based on the person's facial textures and shape (Wikipedia, 2019). This technology is typically used as access control in security systems. Facebook uses facial recognition to detect faces in photos and look for recognizable patterns to identify individuals.

4. IT Management

Virtual agents and chatbots (virtual assistants) are being used from IT support perspective, i.e. to answer user questions. These are business applications that deliver rich information and provides answers to common questions. These technologies can be used for assistance in IT Service Management alongside the Service Desk. Artificial intelligence for IT Operations (AIOps) platforms are beginning to emerge. These platforms combine big data, AI/machine learning and other technologies to support all primary IT operations functions. The data and insights can be useful to automate incident assignment and to enable self-healing incidents. Companies like ServiceNow, are adding machine learning techniques to their offerings to make them more intelligent.

5. Supply chain

AI technologies can be extremely effective in the supply chain. Traditional optimization technologies that helped to predict inventory levels and avoid out-of-stocks, for example, are being replaced by machine learning systems that can continuously monitor sales, weather, and responses to marketing promotions to adjust supply chains. UPS uses AI to create the most efficient routes for its fleet. This enables timely delivery and efficiency.

AI Adoption

Integrating new technologies, such as cloud computing, blockchain etc. with legacy IT systems and processes is always a challenge, AI is not an exception. Given the complex nature of cognitive technologies, AI may be harder to integrate. Though using RPA is relatively seen as easy compared to the rest, i.e. Deep learning and machine learning etc.

With every technology innovation, there are early adopters. According to Gartner, taking advantage of innovations in digital technologies to create new business models, services and experiences remains a priority for many organizations. Gartner uses Hype Cycle to articulate emerging technologies from early innovation to mainstream. Digital native companies, i.e. Google, Amazon, Facebook etc. are known as early adopters of emerging technologies. The rest of the companies, i.e. startups and enterprises usually follow suite. These digital native companies are not only the early adopters but the world leaders using these technologies to their advantage.

Emerging technology adoption should be part of business strategy and organizations should adopt these technologies based on their business needs and not to show off. Though AI is still advancing, it is a revolutionary technology that can offer many organizations with endless possibilities.


© 2017 - 2020 Buhlebenkosi  Consulting 

All rights reserved 

Created with Wix.com