Catalyte Uses AI to Hire Better Software Engineers
Way back in 2004 when nanotechnology became a trendy topic among investors, everyone wanted to get in on the action. It’s kind of like how blockchain and cryptocurrency are today. We’re seeing numerous examples of companies that are changing their names only to see their share prices skyrocket. A great example of this was the recent case of Kodak. Since ICOs in their present form have no equity associated with them, they largely serve as something that speculators can tool around with and that suckers can “invest in”. Contrast that with technology that has real potential like artificial intelligence. Of course we’re now seeing companies that are jumping on the “AI bandwagon” to see their share prices react accordingly. The key takeaway for investors here is that you must approach every claim of “AI” by asking one question. What efficiencies can you demonstrate today that couldn’t be accomplished using traditional software development?
Having a measure of success is key. Companies like Afiniti offer their service for free knowing that they will save their clients so much money that just getting a piece of those savings will actually net them more revenues than if they charged a subscription for their service. With more than 3,000 AI startups operating today, you cannot come to the table with an idea. You have to come to the table with results. We’ve talked before about using AI to improve the hiring process. One such company we came across recently is Catalyte.
Founded in 2000, Maryland startup Catalyte has taken in $38.7 million in funding to “use artificial intelligence to identify individuals, regardless of background, who have the innate potential and cognitive ability to be great software developers“. Their most recent funding round of $27 million closed just last week when they acquired another startup called Surge (more on this in a bit). They’re working with an impressive list of clients that include names like Microsoft, AT&T, Nike, and Geico just to name a few. So what is it exactly that they do?
In a nutshell, Catalyte discovers people who have the aptitude to become software developers and then trains them to do just that. Intelligence is simply the ability to learn, and Catalyte believes that if they can find people who have the ability and desire to learn, they can then be trained in 3-5 months to be software developers. While there is no shortage of aggressive individuals to be found outside the borders of the United States, Catalyte focuses on “on-shoring” as opposed to “off-shoring”. That means that they focus on employing people who live in the United States. The use of artificial intelligence is on the front-end where their “proprietary algorithms” help select high-performing individuals:
We actually gave this a go and had one of our MBAs bury his ego for 60 minutes and apply for a job at Catalyte. While we’re not at liberty to provide a detailed description of the process, we can assure you that it has sufficient rigor to weed out anyone who isn’t really serious about getting a job there. While it’s true that you don’t need a resume to apply, you will still need to provide all the standard information that any job would require such as prior work history, education, references, etc. You’ll then have to undergo about a 2-hour screening process that will look at things like your ability to do math (nothing too heavy), your reading comprehension skills, and so on. After looking at the screening process, we can only assume that they’re using artificial intelligence to refine the process by looking at how the high-performers engaged with the screening process and then adjusting it as time goes on. We’ll talk about how they measure “high performance” in a minute, but first let’s look at what the employee gets out of the whole thing – aside from a job to pay the bills which most people in Mumbai would kill for.
From an employee’s perspective, Catalyte states their value proposition as follows:
Catalyte provides you with training valued at $25,000 (for the entry-level computer programmer program) and $12,500 (for the entry-level analyst program) and (assuming successful training completion) you are eligible for a job. Pay is a minimum of $15 per hour, including health benefits. In return, you commit yourself to success, first in our training and then on the job. By working for Catalyte for at least two years, you receive valuable training for free.
83% of their alumni who complete the program stay in technology, with many landing opportunities at companies such as Amazon, Nike and Microsoft.
Let’s get back to talking about Catalyte’s latest acquisition of a startup called Surge. Founded in 2007, Bellevue, Washington based startup Surge had taken in an undisclosed amount of money to build an onshore “full-service software development and consulting firm” that was on the Inc. 5000 list of America’s fastest growing companies for six straight years. Their value proposition entails offering software engineers at 30-50% lower than market rates while using agile software development methodologies just like Catalyte. (Those costs are probably competitive with John in Mumbai now given that one of our MBAs who worked with offshore teams in Mumbai saw inflation of 9-11% for each of the past three years.) This plug-in acquisition means that the combined company will be one of the largest onshore development providers in the United States.
One thing that Catalyte brings to the table is their ability to measure “success”. Remember we talked earlier about how you might identify “high performing” team members? One way to do this in software development is to look at things like defect rates, price, and productivity. These are all measurable when you use a consistent software development methodology across teams. Both Catalyte and Surge use an “agile software development methodology” which means they can easily begin using the same measures of success across both companies, metrics such as the ones seen below:
Anyone who has developed software for a living can relate to some of the metrics seen above. If these metrics are accurate and consistent over time and across teams, then it’s hard to argue that what they’re doing would not be considered successful when it comes to building better software. It’s these metrics that help make a compelling case for their solution. The same can’t be said for some of the other startups dabbling in this space which claim to be “making the resume obsolete”.
Catalyte is one of the few “we’re using AI to reinvent the hiring process” companies out there that is able to describe the success they see in hiring using objective metrics. That’s more than we can say for companies like Pymetrics which uses “neuroscience games and bias-free AI to predictively match people with jobs where they’ll perform at the highest levels“. Founded in 2013, New York startup Pymetrics has taken in $16.6 million “to reinvent the way companies recruit, hire, and retain talent” by using games instead of resumes. We reached out to Pymetrics to ask about what science they base their games on – twice – and didn’t hear a word back from them. That didn’t really matter though because all we needed to do is look at their white paper on “gender equality” where they actually have the audacity to say that women don’t do well on standardized tests so we should be using games instead.
The entire premise of companies like Pymetrics, is that success should be measured by the race or gender of a team’s composition. Any hiring manager worth their salt knows how dangerous this approach is. For every single role, the population of qualified candidates will determine what the team makeup is, and in many cases, it may not match at all. Trying to force feed people into a role based on their physical appearances is discrimination in its worst form. Using games where “there is no wrong answer” may be a good way to find candidates to work in fashion, fast food, or human resources, but not roles that require technical competencies.
There are certainly many other startups out there that claim to be using AI to improve the hiring process but not all will be using artificial intelligence in a way that demonstrates measurable improvements in the quality of employees that are hired. It all comes down to being able to measure quality. In software development, there are plenty of ways to do that. While we’re not entirely convinced that what Catalyte is doing requires AI, we’re certainly impressed with the outcome. If by building better software development teams they happen to be creating more diverse teams of people with varying life experiences, then that’s even better.