Payment Required

Hunter

Maker

Maker

Maker

Maker

Maker

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

Commenter

@parekh_tanmay thanks! What's exciting about Magnet to me is that it's a 10X better experience for *both* companies and engineers. From a company's standpoint, you just get a queue of strong engineers who want to speak to you (the thing you want as a recruiter/hiring manager). Basically, no one had good enough data in the past to build this (and so you saw "sourcing" tools to help with the work of sending hundreds of outbound messages)

@florin_biro Definitely would be interested in knowing how they assess the engineers have in fact the skills they list

@florin_biro Our technical assessments (introduction here, more detailed breakdown by the engineer who builds it here) are the first-line defense.

We have ~30 assessments on various technical topics. Each assessment is calibrated on a pool of somewhere between ~500 and ~20,000 engineers that have taken it in the past. We also validate those results against the ATS data of companies who use our assessments to screen their own candidate pool. That data shows that high scorers are about 4x as likely to ultimately be hired than replacement-level candidates, so we're pretty confident our assessments are effective.

We also do some extra work behind the scenes to identify and penalize or boost candidates we think are better or worse than the data they enter would suggest. Most of those tweaks are pretty in the weeds, but as an example, candidates who get rejected by employers much more often than they "should" be are deprioritized by Magnet's candidate-selection algorithm.

@jenny_abotinsky Our candidate pool already contains about 200,000 engineers who have signed up for earlier iterations of our product. So in a sense we've *already* sourced the talent, and at present, candidate volume has generally not been the bottleneck for Magnet users. (Typically, if users see weaker results, it's because they're looking for something specific that we haven't built search infrastructure for. This infrastructure is currently a big chunk of our new development.)

In terms of where we're sourcing candidates from today: our largest single sourcing channel is our Screen product. Screen lets companies use our assessments to screen their own applicant pools to (a) cut down on fluff and (b) identify hidden gems they would otherwise have missed. Screen company users send our assessments to their candidates, and once candidates finish the assessment, we prompt them to sign up for Triplebyte to use their results with other companies in the future. We add a few thousand candidates each week through this and other channels on top of the existing 200k.