A simple google search of “Digital Transformation & Artificial Intelligence” returns millions of results. In current usage, Artificial Intelligence (AI) covers a range of topics – Machine Translation, Deep Learning, Social Analytics, IOT etc. How should you leverage AI in your organisation? By diving into this Digital pond at the right place – along with a judicious mix of People, Process and Technology (PPT).

Technology

Algorithms have long ceased to be the preserve of Mathematicians and Computer Scientists. A Procter & Gamble backed start-up is touting a Digital Skin care device with a digital camera that captures 24000 skin pictures per use, analyses the data with a 70000 line algorithm and then leverages micro-fluidics to deliver make-up (cream?) with 120 thin thermal inkjet nozzles depositing 1,000 Optimizing Serum picoliter droplets to transform the look of the skin. WOW!

The Technology challenge is multi-fold:

1. Store and readily access enormous quantity of data leveraging cloud data stores (In the above example, do the math for millions of consumers)
2. Preserve the context of the event (e.g. Skin Pictures, algorithm version and the fluid application for each customer every time needs to be recorded because all those variables are specific to each context)
3. Integrate with other CRM applications in the organisation.

Process

Transaction systems have gradually evolved over the last 40 years or so. From generic payroll and procurement systems to industry specific (e.g. Pharma) to sub-industry / geography focused systems (e.g. Pharma formulation Contract Manufacturers in North America). Users expect such out-of-the-box systems to reflect the “Best practices of processes” and that they would just provide the organisation specific data and use the system for years; often significant investments are justified with a 3-5 year ROI. This Process-Data paradigm is significantly different in the AI world as they dynamically drive each other.

To address the Process Challenges,

1. Look at all data, within the organisation and outside that needs to be harnessed (Do remember that when you start the project, you will not know all the data that will be required!)
2. Develop algorithms to glean meaningful insight
3. Implement changes in processes based on the insight

These three steps will be iterated as business outcomes are assessed and improvements expected. Users will demand more sophistication (Read Complexity!) in the processes which, will further increase the Technology Challenge.

People

Given the Process and Technology challenges, classical project team structures would flounder.

To address the People Challenges

1. Approach AI initiatives using the Japanese Kaizen method of Continuous Improvement (This is critical to sustain enthusiasm and commitment beyond the inevitable initial hype!)
2. Leverage Communications to reiterate that Technology is just an enabler for an end-goal (In the P&G example, the user is looking for flawless skin!)
3. Manage proactively the egos and team dynamics of algorithm spouting PhDs, brand flaunting marketers and Cloud Benefits touting IT service providers

Summary

A successful AI initiative is the outcome of a well-balanced tripod standing on the legs of the appropriate process leveraging the required technology with the right people.