Launch HN: Searchlight (YC W19) – Hiring based on past performance, not resumes

Hi HN community! We’re Anna and Kerry, co-founders of Searchlight (https://www.searchlight.ai https://www.searchlight.ai). Our software helps candidates be judged by their past performance rather than their resume or where they went to school. We built this product to help job candidates and hiring managers. With platforms like Linkedin and Indeed, hundreds of applicants with indistinguishable resumes apply for the same job with just one click. Kerry and I both have backgrounds in software engineering, and we were frustrated by how time-strapped hiring managers increasingly over-index on the “snob test” (a.k.a. where the candidate went to school) or contrived technical screens [1][2]. We’re also twin sisters who went to the same school and worked at the same companies. We look indistinguishable on paper, so we are especially keen to bring a new product to the hiring space that will allow candidates to express their individuality beyond their resumes. When we looked at the landscape of current hiring tools, we realized that the majority of them are self-promotional (resumes, personal websites, Linkedin, etc) and difficult to substantiate at first glance. This disadvantages people who aren't good at promoting themselves, or don't like to, and these are often the best candidates! We saw that a poorly conducted technical screen can penalize the most talented engineers. Worse yet, we learned that take-home coding challenges are a real pain point for certain demographics, like parents who don't have the time to thoroughly attack a 24 hour coding challenge because they have to take care of their kids. This made us think - why are we ignoring the the perspectives of people who actually know what it's like to work with a candidate? This data is the most indicative of success on the job [3][4], but isn't currently being leveraged until the end of the process, if the employer conducts reference checks. This is why we built Searchlight to better assess candidates early in the hiring process. Currently, we work directly with employers to invite their applicants to the platform. Job seekers can invite as many advocates as they want to speak to their accomplishments and capabilities (some invite as many as 10). The references share feedback like specific examples of how the candidate demonstrated desired competencies and how future managers can set the candidate up for success. Then, we analyze this feedback to assess candidate-position compatibility by matching the requirements of the role to the candidate's strengths. Our recommendations for strong candidates are based on a mix of quantitative factors like average ratings of core competencies, and qualitative factors like work style and environmental fit (which we currently human QA). One of our core beliefs is that every candidate is exceptional in their ideal environment, so all the feedback gathered on Searchlight - regardless of whether the candidate gets an offer - is saved and available for the candidate to use and share. We aim to make the hiring process more fair. We are building trust and legitimacy into our platform by tying each reference to a specific job experience, verifying references through work emails or Linkedin profiles, and keeping the feedback hidden from candidates. While no tool is perfect, we know that the insights surfaced by Searchlight allow for better decision-making than traditional resume scans, with no extra time commitment for employers. We are especially excited to see that Searchlight is already helping diverse applicants get to the on-site interview stage after being initially screened out. We'd love to hear about your experiences in today's hiring process and if Searchlight would be helpful to you! Thanks for reading. [1] https://news.ycombinator.com/item?id=15688972 https://news.ycombinator.com/item?id=15688972 [2] https://news.ycombinator.com/item?id=2175147 https://news.ycombinator.com/item?id=2175147 [3] https://dornsife.usc.edu/assets/sites/208/docs/Ouellette.Wood.1998.pdf https://dornsife.usc.edu/assets/sites/208/docs/Ouellette.Woo... [4] https://www.linkedin.com/pulse/how-predict-on-the-job-performance-30-minutes-lou-adler/ https://www.linkedin.com/pulse/how-predict-on-the-job-perfor...