Juicebox (PeopleGPT): AI in recruiting

In our category "HR Tech in US" we present HR tech companies from the USA and what the European HR world can learn from them.
In this interview, I spoke with David Paffenholz. He is Founder of Juiceboxa company that offers an AI solution for recruiting.

David, you developed PeopleGPT with Juicebox to bring artificial intelligence (AI) into recruiting. - How does your approach work and what makes it special?

Imagine if you had a recruiting assistant who's read through hundreds of millions of profiles and built an understanding of each individual's experiences, skills, and interests. You can ask the assistant to help you set up searches for best-fit candidates, craft personalized messages for them, and present you with an organized list of results.

That's what PeopleGPT brings to market - today - an AI powered people search engine. You enter a description of who you're searching for, and PeopleGPT will 1) find best fit profiles across 30+ data sources, 2) write personalized outreach, and 3) export those candidates, including candidate summaries, to your ATS / CRM.

Will AI put recruiters out of work?

No, but it will transform their jobs. 

Sourcing will evolve from reviewing thousands of profiles to selecting among best-fit candidates that the AI found. Candidate relationships will become even more important, as personalized messaging becomes commonplace. Employer branding will take on increased importance, even involving role-specific branding initiatives.

Overall, more time will be spent in AI-assisted workflows, guiding software to help achieve goals. It will result in more impactful TA departments, with more attention placed on candidate experience, higher conversion rates across all funnel steps, and greater hire success rate.

How do you envision the future of your AI? Do you think beyond the recruiting sector?

We're focused on recruiting. 

That involves 1) adding more data sources into our platform, many of which have entirely untapped data on candidate skills and experiences, 2) expanding our AI features, including entirely new workflows, and 3) deepening our integrations with the 60+ ATS and CRM platforms we currently support.

You know the German and American markets. What do American tech companies do differently to German ones?

I'm more familiar with the American than the German market, but some of the key differences we've observed (purely anecdotal):

  • Active Sourcing: more commonplace across roles in the US
  • Candidate Outreach: outreach still leans towards LinkedIn or Xing in European, though there is increased willingness to try other channels (email, phone, WhatsApp)
  • Personalization: generally high personalization and strong engagement in the European market

Why did you set up in the USA? 

I first moved to the US to go to college and have stayed here since. In my mind, it's undoubtedly the best place to build a tech startup: you have a high concentration of like-minded founders, access to capital, and a large talent pool with previous startup experience. We started building PeopleGPT while going through Y Combinator, a startup accelerator in San Francisco, and it was an easy decision to stay here since.


David, thank you for the interview and your insights!


HR-Tech-in-US-Juicebox

AI in recruiting: How is your job changing?

No matter where you look, whether it's on TV, on LinkedIn, in the newspaper, everywhere you look people are talking about AI and how it's going to change everything. But how realistic is that really? Where is AI already finding its way into our everyday jobs today? And how will your recruiting job change as AI takes over more and more tasks? That's the subject of this article.

What is artificial intelligence (AI)?

Artificial intelligence (AI) is a technical application that aims to digitally emulate human intelligence. The aim is to enable machines and programs to adopt and improve human thinking and behavior and to imitate them in various tasks. 

AI is based on huge amounts of data, which are put into relation with each other via algorithms. The AI is trained to recognize patterns and correlations between the data and to draw conclusions from them. The goal is usually for the AI to be able to learn independently and improve itself.

There is almost no area in which AI has not yet found its way or will not do so. Whether it's finance, medicine, production, communication, AI is everywhere. Without you always noticing or questioning it. For example, the facial or fingerprint recognition of your smartphone, digital assistants such as Siri or Alexa, autonomous driving, recommendations on Spotify or in social media, or even your navigation system. The list is already endless today. And it's getting longer all the time.

What advantages does AI bring?

AI is the logical consequence of what we humans have always done: We make things easier for ourselves. It's easier (and sometimes safer) to use your fingerprint to unlock your smartphone than to have to type in a four- to six-digit code every time. It's convenient to learn about new artists you might probably like on Spotify that you might otherwise miss. And it certainly equalizes traffic already when many people use Google Maps as a navigation system and bypass the red areas directly.

AI makes the world simpler, cheaper and faster. And it opens up potential that would probably never be possible with human power. Because it is more powerful and always available. It doesn't need to rest and doesn't make mistakes. It does what people tell it to do, without any will of its own. And with sufficient data, it can make predictions that humans cannot make validly. 

That already sounds very positive. But where there are advantages, there are also disadvantages.

What are the disadvantages of AI?

When it comes to AI, one of the main points of criticism is ethics. After all, the technical application is still based on human creativity, which develops the AI. According to some researchers, this cannot be completely free of prejudice. Furthermore, there is the question of responsibility. Who is responsible if the AI should make a momentous decision? For example, if a human is killed in a car accident with a self-driving car. Is the manufacturer of the car liable? Or the developer? And would anyone then even develop such an AI if they knew they would be liable? There are many new and unresolved issues here that we need to address.

AI continues to bring the risk of unemployment in many areas. Jobs are very likely to be replaced or changed by AI. But it will also create many new professions that we don't even know about today. 

Other disadvantages of AI are its lack of empathy and its unpredictability in the case of self-learning AI. In addition, it is important to remember that the AIs that have been popular so far, such as ChatGPT, are language models. That is, these AIs have been trained to learn, understand, and reproduce language. This does not yet mean that all the information that this AI reproduces is correct. Thus, such an AI can sound convincing, but tell plain wrong.

Where can we already find AI in recruiting today?

The world is changing. AI is already here in many areas of life and continues to make inroads, and we can already find AI applications in recruiting. They relieve recruiters and budgets, optimize processes and help to achieve KPIs. The question is: In which recruiting areas can we already find AI today?

Assessment Center

Assessment centers can be supported by AI, for example. Certain tasks can be used to test the competencies and skills of a candidate. These are automatically evaluated and compared. Both in advance, when selecting who is invited to the assessment center, and during the process. In addition, speech analysis tools can be used to obtain information about the candidate's personality and behavior.

Applicant analysis

Voice and video analysis tools are also finding their way into applicant analysis. The entire process, from the analysis of application documents to the evaluation of interviews, can be accompanied by AI. 

Chatbots

The chatbot has already become a "classic". With this, website visitors to the career site can get in touch with the company via a chat without the need for direct action by an employee. The chatbot can respond to common questions and forward them to an employee if necessary.

Interviews

Today, it is also already possible to have entire interviews, at least digital initial interviews, conducted via AI. This can evaluate language, choice of words and facial expressions and deduce whether the candidate is a good fit for the company.

Job postings

AI can also support the creation of job ads. By analyzing job ads, job categories, keywords and possible matching job boards can be derived. In this way, AI supports the smooth retrieval of the job advertisement for the applicant.

How is AI changing recruiting?

You will therefore already find AI in the entire recruiting process. As a rule, however, it is more of a support than a fixed component with its own package of tasks. As a rule, there is still a certain susceptibility to errors. But this development will also continue. Errors will be corrected, systems corrected and further developed. AI will become an integral part of the recruiting process and increasingly take over tasks independently. Above all, recurring tasks such as applicant screening and matching (in the Active sourcing), will (be) taken over by AI. 

This leaves the recruiter much more time for the essentials. Hiring the right employees. The bottom line is that this decision will remain one that is not made by AI. Who will be hired when the company has to decide between two or more candidates? Who decides in favor of the candidate who is "worse" on paper, even sometimes? That's what humans do. Because he has one thing that AI cannot reproduce: Gut feeling. AI will therefore change a lot, but not everything. In the end, people still work with people. It doesn't matter what the other person's high school diploma is or whether he or she has frowned in a conversation. Only people and their gut feeling know how to evaluate that.


Tech Recruiting

FAQ

AI Recruiting: What is it?

AI Recruiting (Artificial Intelligence Recruiting) is the technical support of the recruiting process by means of software and automation. 

How is AI Recruiting changing recruitment?

AI Recruiting (Artificial Intelligence Recruiting) is now finding its way into the entire recruitment process. This can be supported by AI from the initial contact of the applicant with the career site, for example through chatbots, to entire interviews.

How is AI changing human resources?

Artificial intelligence (AI) already supports many areas in human resources. For example, in recruitment, personnel development and employer branding. This article shows you how AI is finding its way into recruiting.

The future of software development: Where is the journey heading?

On May 12, 1941, Konrad Zuse presented the Z3 in Berlin as the world's first computer. The Internet, then called Arpanet (Advanced Research Projects Agency Network), has been picking up speed since 1969. In 1994, the first search engines were launched with Yahoo and Lycos. In 2004, social networks such as Facebook and XING appeared on the market. And since Steve Jobs presented the first iPhone in 2007, the mobile Internet has become indispensable. 

Many still know the world without smartphones. The time when you still had to click keys several times to type a letter into an SMS. That world was only 15 years ago. Today, we talk to our devices and the text is created automatically. Our world has been "software-ized.

1. The importance of software development

Modern life would be inconceivable without software. Consciously or unconsciously, we constantly encounter software in our everyday lives. The preparation of our morning coffee, the traffic lights on the way to work, the organization of supply chains so that we can buy our groceries in the supermarket. Software is in use everywhere. And it is becoming more and more effective and faster. It facilitates our lives enormously and we have all developed a small or large demand for it. Who hasn't been annoyed when the website loads too slowly or the registration on a website was answered with an error. 

To meet this demand and further softwareize our world, software development is continuously evolving. Sometimes so fast that even developers have difficulty keeping up. 

2. The future of software development

Programming languages, frameworks, libraries: technology is evolving ever faster. Frontend-For example, frameworks are constantly being updated, with major breakthroughs also occurring. What is "hot shit" today will be replaced by more efficient technology in a few years. This is what is currently happening, for example, with the programming language Rust, which in certain Big Data and machine learning applications. Python replaces. 

But what does this mean for the future? How will software development change? And what does that mean for us as tech recruiters as a result?

The future is automated: Artificial Intelligence (AI) in Software Development

One of the big topics of the future is AI. The potential of AI is far from being tapped. Just as in every other area of life, it will continue to make its way into software development. After all, software development is not just about ingenious virtuosity, but often quite routine work. And this is what AI can take over. For example, AI is already being used in test automation to help ensure software quality. Added to this is its application in the practical day-to-day work of developers. The automatic writing of code has already begun.

In the future, AI can be used to implement very interesting ideas. Let's take Natural Language Processing (NLP) as an example. As the name suggests, this involves the machine processing of natural languages such as German. We are all familiar with this from voice assistants such as Siri, Alexa and Co. We say something, the machine understands it and responds. If we transfer this mode of operation to an AI that could write software on its own, we can imagine a software developer sitting in front of his computer and dictating code as a first step. That would be conceivable, but not very exciting. In a second step, however, we could also imagine ourselves, as non-developers, sitting in front of the computer and describing our ideas of the program, which the AI then translates into a working program. In this way, everyone could talk to the AI as if it were a software developer, except that the AI could potentially take over the work of an entire team of developers. 

The future is simple: low code and no code

What can already be found today, especially in website construction kits, will be found more and more in the future - low-code and no-code. As the names suggest, it's all about building programs with little or no code of your own. Essentially, individual building blocks are simply put together using drag & drop. That goes fast and can be made also by Nicht-Entwicklern. There are limits to the amount of customization that can be achieved. In the future, however, more and more people will be able to participate in the softwareization of the world, even if they are not developers. 

The future is secure: DevSecOps - security already in the process

The more creative we become in developing software, the more creative the group that attacks that software becomes; namely, hackers. According to the industry association, cyber attacks cost the German economy Bitkom 220 billion euros per year. Almost every tenth company even sees its business existence threatened by cyber attacks.

The topic of IT security is central. For this reason, it will become more central to the development process in the future. With the introduction of DevOps, two departments (Development and Operations) have practically been linked over the last ten years. Modern developers therefore not only write their code, but can also bring it live themselves. Previously, this required a different IT specialist (Operations), who only took care of the infrastructure required for this. Merging the two issues therefore brings more speed and agility to the development process. However, the security aspect is all too often neglected. Seen in this light, they are still located in a different department. So now DevOps and security are being brought together and DevSecOps is being created. This means that security issues are considered directly in the development process. 

The future is mobile: Native Mobile

In our modern society, smartphones and tablets take up a significant part of our time. For this reason, a mobile-first approach has already become established in the past. An application is therefore primarily developed as an app for the smartphone or tablet. In development, this goes hand in hand with the fundamental question: which end device should be developed for? Android or iOS? Special programming languages have already been established for this purpose, which improve the speed, memory security and parallelism of the apps. These include Swift and Kotlin, for example. 

3. The future of tech recruiting

What can we derive from the future trends of software development for tech recruiting? 

Due to the mobile-first approach, a higher demand for mobile developers can be expected. Apart from the existing ones, new developer jobs will also be created. One only speaks of a DevSecOps approach, for example, when a separate security team is really superfluous. Usually, a "security champion" is appointed within the development team who, trained in security topics, assumes responsibility for security fixing. As a result, the profiles being sought are becoming more extensive in their requirements. DevOps developers need security expertise. Security specialists go more into the development process. New role titles may emerge that we don't even know today.

As fast as the technology world is changing for software developers, it is also changing for us tech recruiters. An understanding of software development, languages, frameworks and the like is essential for successful recruitment. For this reason, tech recruiters will also have to deal intensively with the contexts and technical language in the future. 

As artificial intelligence finds its way more and more into software development, it is important to observe how teams change as a result. Routine work or simple tasks are usually taken over by junior developers. Whether these will then become "superfluous" and be replaced by AI remains to be seen. 

The list of possible trends is not exhaustive. It is notoriously difficult to predict how the future will shape up. In any case, there are no limits to the ideas. What is your idea of the future of software development? Have you ever talked about it with someone in your department?


The future of software development

FAQ

What role does AI play in software development?

Artificial intelligence (AI) will increasingly find its way into software development. It is already being used today in test automation to support the assurance of software quality. The automatic writing of code has already begun.

What is DevSecOps?

DevSecOps is a word combination of Development (Dev), Security (Sec) and Operations (Ops). DevSecOps is an approach to software development and deployment that considers security at every step of development.

What does the future of tech recruiting look like?

Understanding software development, programming languages, frameworks and the like is essential for tech recruiters. Therefore, they will have to deal intensively with the contexts and technical language in the future as well. With a greater need for IT talent, more tech recruiters will also be needed to find and hire them.