AI technology can transform a company in many ways. IT can help reduce costs, increase revenue and give your company blue water from your opposition. In most cases, you may already have an idea of an AI project you want to develop.
A good start is by looking at your business pain points and the opportunities. You may unearth key ideas. Some developers look at issues or problems from customers or staff perspective and use these as a starting point. Try to brainstorm with key people in a business and you should unearth some real gems.
What you want to achieve?
Work out the purpose or end goal. Make sure that it is something attainable and that the whole company can adopt.
Start small with quick wins
As Andrew Ng1 always points out – ‘start small and go for quick wins’. There is huge merit in this. If you take a big idea and break it down to small wins or gains, your team will see progression and a return on investment. We are not all like a heavyweight digital companies, with bottomless budgets and time. We do work under realistic constraints.
Work out your constraints
The best thing to do is work out what you have and what need for the best results. The keys things to work out are:
- Data needed – Look at your data, do you have enough, is it clean?
- Costs – What are the direct and indirect costs?
- Time – How quickly can you achieve your quick wins and your longer term goals?
- Resources – How much resources do you need to achieve your goal?
- Saves you the most amount
- That gives you the greater return
- Adds equity to the company for the longer term
Compare and contrast
Run each project through these needs and returns and compare the results. Try and pick a project that has little impact, but has the greatest return. The idea is to develop a proof of concept, win your company over to the bigger picture.
Layout your AI strategy or plan
Once you are armed with pain points, the variables, constraints and returns you can then work out a schedule. Keep making sure that through the development, that it does not slide and that hits the goals with the greatest ROI. There are several techniques for a successful plan which will be discussed in future articles.
If you are unsure, consult me or use The AI finder to get the best consultant for your needs. What you want to do is have a success rather than a white elephant and red faces. This is a brief rundown of what is needed, but I can give you a more deeper plan of attack if needed.
Copyright Mike Nash – If used, please mention my name and link back to this page.
1) Andrew Ng - Leading AI computer scientist. Once head of Google Brain now working for Landing AI.
Keywords in article to help search
ai, ai projects, ai planning, ai strategy,