” alt=”” aria-hidden=”true” />
Image credit: Maris Vector/Getty Images
Companies already have a wealth of tools at their disposal for headhunting technical talent, but new startups are bringing together the worlds of GitHub and LinkedIn to create a database of the best candidates for a particular software development. With that in mind, I want to step into recruiters. It does its job by using AI to “guess” skills from the code you write.
program AI: A company called GitHub allows recruiters to search for developers based on their technical skills, libraries used, or simply contributions to projects on GitHub.
Founded in San Francisco in 2022, Prog.AI is the brainchild of a CEO, Maria Grineva, WHO (I sold my previous data startup). We invited Orb Intelligence to Dun & Bradstreet in 2020. CTO Fedor Soprnov, formerly a machine learning researcher at Russian tech giant Yandex and product head Dmitry Pianov has worked in product teams for companies such as Yandex and Replika.
The company initially focused on hiring, and this week rolled out its first product for recruiters in closed beta. Grineva sees a wide range of use cases beyond just helping companies take on a technical role. This includes facilitating developer relationships, such as inviting developers to join our community or contributing to open-source projects. Requires expertise on specific issues. You can even help development tool companies market their products.
“This week we are launching Prog.AI for tech recruiters. Extend your SaaS with Prog.AI, learn about your existing developer community and reach your target audience,” Grineva told TechCrunch.
Prog.ai today announced that it has raised $1 million in pre-seed funding from a Germany-based angel fund to accelerate its commercialization: Angel investment, Brooklyn Bridge Ventures and numerous angel backers, including one of Spotify’s first employees and former CTO Andreas En.
So how does Prog.ai actually infer skills from public source code? Git clone command creates copies of millions of public repositories and branches. Prog.ai then analyzes each git commit, inspecting the code snippet, file path, and commit subject to figure out what it is.
“For a given project, you can see who is the core architect, who develops the backend or frontend, who focuses on UI/UX, who does QA and testing, and who is the technical writer.” says Grineva.
Prog.ai also delves into git actions such as pull requests (decline and approve, comments, open issues, etc.). This helps Prog.ai “understand” different roles and levels of involvement of project contributors.
“We handle not only well-known open source projects, but also ‘pet’ projects, tests, forks, and even Coursera and Udemy training projects that our engineers publish on GitHub,” Grineva added. “Overall, we process about 1 billion commits per year on GitHub to get a very accurate profile of every engineer’s skills.”
Under the hood, Prog.ai relies on OpenAI GPT. For example, we tune language models touted in high-profile open-source projects and StackOverflow articles to help derive code quality scores.
Prog.ai users can create a list of top experts in a particular area, such as ‘Large Language Models’ or ‘Computer Vision’, and generate a leaderboard of top performers in any area. Alternatively, you can submit a list of repositories and create a ranking of all contributors by the number of commits they have made.
Effectively, recruiters and employers can tailor their searches to the parameters they want, such as skill areas, programming languages, years of experience, and more.
But understanding code is only part of what Prog.ai has to offer. The main selling point for recruiters is connection To that end, Prog.ai has a built-in email outreach engine and utilizes a sales engagement platform: Reply.io.
“Users use our search to create a list of relevant candidates. They then create personalized email sequences, mention candidates by name, browse projects, and I can explain why I think the position is right for me,” says Grineva. she said.
Recruiters will want a more comprehensive view of a developer’s skills, education, and work history that they probably can’t get from GitHub. This is where LinkedIn joins the fray, Prog.ai collects publicly available data and coordinates with corresponding individuals on GitHub, and this is the platform’s special source, she says, Grineva. said. By meshing data from two of her widely used platforms, we can build a more granular picture of potential candidates.
“I think having a GitHub and LinkedIn profile adds a lot of value because engineers are typically not very good at promoting themselves and often don’t even have a full LinkedIn profile.” says Grineva. “Furthermore, on LinkedIn, people introduce themselves, which means the information is subjective. Inferring an engineer’s skills not only eliminates subjectivity, but also means that companies can evaluate candidates uniformly.”
Of course, none of these offer a perfect recruitment path. Bringing together two huge and disparate datasets is no small feat. Profiles with similar names and histories are more likely to be confused, which increases the chances of errors. This assumes the person has a LinkedIn profile in the first place, but it’s possible that they definitely don’t. Internally, however, Grineva says steps are being taken to address at least some of these potential pitfalls.
“Matching two large datasets is not an easy task because the information people make available on GitHub is sparse and many engineers choose to remain anonymous on GitHub. because there is,” explains Grineva. “We built our own fuzzy matching system that takes into account not only his name, username, email address, but also his place of work, expertise, interests, etc.”
Additionally, Grineva said computer vision will be used to compare profile avatars across platforms. This is not foolproof by itself, but serves as an additional tool along with other verification mechanisms.
As of this writing, Prog.ai claims to have contact information from approximately 70% of all profiles in its database. This clearly means that 30% are missing that important data. He said that typical use cases don’t always revolve around reaching out.
“Another important use case is data enrichment,” she says. “Customers can search for a full candidate profile by GitHub handle, her URL on LinkedIn, or contact email. In this case, we can only match her 70% who have an email. increase.”
There is also a huge elephant in this room. Doesn’t Prog.ai simply encourage “cold callers” trying to contact developers en masse?
“It’s risky, but it’s important to first realize that recruiters are already trying to reach out to developers. This is now being done through other tools. Some technologies Recruiters manually extract contact information directly from GitHub,” says Grineva. “That said, recruiters currently do this with bad or limited insight into the developers they are reaching out to. Opportunities are not suitable for developers. As a result, these emails are treated as spam.”
Regarding recipients of reach-out campaigns powered by Prog.ai, Grineva said the platform is “fully GDPR compliant” and developers can ask to have their profile removed or edited, as well as opt-out entirely. said. email outreach.
show me money
Prog.ai is still in its early days and is experimenting with different plans, but the company essentially operates a SaaS-based subscription model, with pricing based on the number of contacts a user accesses. This starts with “Free” for up to 100 contacts per month, and goes up to the “Recruiter” plan for $530 a month with advanced search features and 3,000 contacts. We also offer custom-priced enterprise plans available upon request.
Also from LinkedIn’s own Talent Solutions products, Zoominfo, SeekOut, TalentOS and Hire EZ. But Grineva says Prog.ai’s focus on purely technical talent and GitHub’s scanning smarts set it apart. By extension, this can It means more targeted headhunting efforts where the goals of recruiters and candidates are more closely aligned.
“As an engineer myself, I get a lot of messages from recruiters that have nothing to do with me, and I see this problem firsthand,” says Grineva. “I think it’s mostly a data quality issue. Recruiters don’t have enough information to match me for interesting opportunities. Our goal is to reduce the noise developers get today. We believe that by giving recruiters better information, both developers and recruiters will benefit.”
DELTA Data Protection & Compliance, Inc. Academy & Consulting – The DELTA NEWS – firstname.lastname@example.org