Harj Taggar, one of Y Combinator’s earliest partners, is very familiar with the pain of hiring for new CEOs.
“When founders start companies, they’re not really sure how to put hiring practices in place. They’re not sure which questions to ask or how to handle technical interviews,” he said.
So while he was at the early-stage firm, he started building out systems to help companies find and bring in engineers, even creating an interviewing checklist that is still used today. Now he and Ammon Bartram and Guillaume Luccisano from SocialCam are taking some of those learnings and starting TripleByte, a recruiting company that specializes in technical hiring.
Their first product is a common application to work at companies in the YC portfolio. The first screen involves answering a few programming questions, then they’ll do a 15-minute technical phone call. TripleByte does not ask for resumes in the first pass, to weed out bias that could disadvantage groups that have the skills but haven’t gone to brand-name schools.
After that, they’ll do a 45-minute technical interview remotely through a screen share. If they pass, they’ll pick up to five YC companies that seem interesting to them. If the YC companies feel that they’re a good fit with the mission and team and extend an offer, TripleByte will help the candidates manage offers with salary and equity data.
If they don’t pass, TripleByte will give them advice on how to improve and they’ll be encouraged to apply again (just the way YC has done with teams that haven’t made it on the first or second try.)
As for their revenue model, TripleByte takes a 25 percent cut of an engineer’s first-year salary, which is a fairly typical model for recruiting agencies. But they’ll offer a six-month guarantee. So if the hire doesn’t work out, TripleByte will refund everything.
“Most recruiting firms will do one month. Sometimes you can negotiate up to three months,” Taggar said. “No one I’ve heard of does six months.”
Their goal over time is to accumulate as much data as possible to understand what makes a successful engineer and how to identify them. It’s similar to what Laszlo Bock has done internally at Google as the company’s senior vice president of people operations. Bock even recently wrote a book on it.
But this is a horizontal technical HR layer that spreads across many companies, instead of being contained inside one enormous corporation. Over time, Taggar hopes they’ll be able to create a more predictive technical interview.
“We’d like to have data across the lifetime of an engineer,” Taggar said. If TripleByte’s software was even trained on a sufficient volume of data, Taggar imagines it could even make the hiring decision if technical ability is the sole criteria.
It could also be a credential that competes with the brand-name seal of approval that attending MIT or Stanford usually confers on an engineer. The cost of tuition for these elite schools, along with all of the other ones that follow them, is rising faster than the pace of inflation. They have immense positive student selection bias because of their selectivity, and tech leaders like Peter Thiel have questioned the value that they offer beyond that network. But there are clearly people who don’t go to these schools who could be successful, so TripleByte intends to identify them. Even TripleByte’s own founders fall into this category. Bartram was homeschooled until 18 and Luccisano grew up and attended university in France.
There’s a few other things that stand out to me about this idea. One, it’s another example of how different elements of companies are being horizontally sliced off. In the way that YC-backed Zenefits handles benefits or Checkr handles background checks for companies with lots of on-demand contractors, TripleByte is trying to own technical ability assessment.
Secondly, it’s an example of how YC is starting to operate as its own economy. A few years ago, Taggar told me that the YC network was starting to be a place where startups could get critical mass for their first products.
But now, the family of startups is so large that even the companies inside YC are a formidable market in their own right. Layering on the ability to funnel in lots of engineers quickly to various YC startups could give the accelerator even more of an advantage (even though Taggar says he wants TripleByte to work for every company.)