How Lytica became a unique analytics company: Part 11
Every now and then it doesn’t hurt to stop and ask the question: what are we doing? I started asking this while the team has been going through verification of our AI work pertaining to our current and future products and services. While I have probably made a semi-infinite number of PowerPoint slides on what Lytica does, I am not sure that any of them really get to the essence of what we are trying to do. This is more than a marketing problem because sometimes, when I explain it to my family, after all of the jargon and acronyms I get blank stares and the question, “But what do you actually do?”
If I think through the products we have made and the ones that we have on the drawing board, they seem to fit into two broad categories – characterization and prediction. Our AI initiatives are focused on creating capability for and improving both.
Our characterization activities center around quantifying the distinctive nature or features of something to do with your supply chain or supply base. As always our focus is on cost, security of supply (risk) and compliance. To date, the best Lytica example of this is competitiveness, which is a characterization of spending effectiveness using market pricing as a reference. We refer to your competitiveness score as the Lytica Competitiveness Index (LCI). We use your LCI as the basis for price prediction but it offers more and, because of this, we are honing our tools and database to increase its precision and value. Additional characterizations apply to component manufacturers.
On the prediction side, we have achieved a favourable position being able to identify which components are properly priced for your company based on your spending characterization as well as identifying components that are outliers. For those mispriced outliers, we recommend a target price that is usually attainable. It’s interesting that some suppliers have trouble with this as they see margins shrinking, yet they are blind to the fact that many of them are pricing components far too aggressively for their market. They should characterize their quotations and analyse their win/loss ratio to strengthen their margins! Players on both sides of a negotiation can benefit from Lytica’s technology.
With the characterization of risk comes the ability to make predictions which underpin some of the products in development for release next year.
So, what are we doing? We are helping people measure how well they are doing with their electronic component materials spending and supply base design which enables them to take steps to improve their position in both spending and risk.
With the Advanced Technology Center, we are trying to do this cheaper, faster and better using technology based on machine learning and AI. These new technologies speed up analysis and enable new insights into what’s driving a client’s results. We are exploring which factors determine a client’s LCI and if we can correlate that LCI with business performance. We have data to show how improvements in one of our customer’s competitiveness over a two-year period led to substantial improvement in earnings per share and ultimately to a successful IPO. Just how general are correlations like this in our business requires investigation, as does determining if AI can be used to go from correlation to causality.
This is all interesting stuff, but reality brings me back to the work of our 2017 program and the complex and detailed verification currently underway. We are seeing the power of our technology increase our productivity in creating new reference data – millions of components at a time, where previous methods would struggle to create hundreds. Our immediate challenge is its verification, particularly when anomalous findings need explanation. Such is R&D, where innovation often follows a crooked path with ups and downs. The good news is that we are making progress and are on track for some preliminary transfers to production this year. The first gains from the 2017 program will be improvements to our Component Cost Estimator product and the new CCE mobile application. These will be followed by advances in FBDC match rates and process times.
We know what we are doing, are progressing and are getting better at explaining where we’re going.
I think my family is actually starting to understand.