Dr Nakeema Stefflbauer wears many hats, but sees a common thread running through all of her different areas of work. “I want the tech industry to live up to what it is supposed to be,” she says.

“That means using tech innovation to improve and transform the society we live in for the better – not just focus on space travel, robotics and cool-sounding virtual experiences.

“That ability to transform society is what hooked me into working in tech. It’s about attacking problems of inequality and access that we find in the offline world.”

Stefflbauer’s founding of FrauenLoop, a German-based social enterprise, is “about the question of who is accepted as a techie and who is considered competent in terms of technical knowledge”.

She says: “It’s the idea of insiders versus outsiders in an increasingly digitised society and addressing that in a tangible way, not just dismissing it as too complex to take on.

“My work as an angel investor at Atomico fits that pattern too. Why are we using tech mainly for convenience services and sidestepping the need for entrepreneurs and innovations that address way more pressing problems faced by people all over the globe?”

Stefflbauer identifies some of these pressing problems as equal pay, women’s health, sustainable food and water, technology platforms, and access to affordable housing and education.

“All these bigger challenges still plague a lot of the globe,” she says. “We’re not going to moonshot our way out of addressing them, so as an investor I’m looking at what kind of innovation exists that needs funding and amplification. I’m seeking visionaries who can re-imagine our future – not just reimagine how we get food delivered.”

Stefflbauer’s motivation largely stems from her sense of being an outsider when she moved from academia into tech.

“Gatekeeping is very prevalent,” she says. “People in tech use insider information, specific language and acronyms to keep out people who aren’t from an engineering/computer science background. They justify it with a culture and discrete set of knowledge that you typically either have or don’t have.

“When I started FrauenLoop, I wanted to help people who felt they could never break into the tech industry because they weren’t ‘the type’. I wanted to ensure women knew success in the industry is very much about motivation, learning mindset, the ability to focus and tenaciously master new skills – not about being a type, playing a role or ‘looking the part’.

“They stopped requiring university degrees for jobs at the biggest US tech companies [Google, Apple, IBM] years ago. Just think about that – some of the most financially successful and technically advanced companies don’t care about your university credentials, so why are we still there in Germany, the UK and France?”

Stefflbauer, an American now based in Berlin, was prompted to create FrauenLoop by a sense she was “going backwards by decades” after arriving in Europe. “I was – frankly – shocked by the sexism, the classism, the racism,” she says. “I wanted a specific change in the ways the tech sector doesn’t live up to the ideal of being an innovative, globalised meritocracy.

“I wanted more people, not just women, but more people from a wider range of backgrounds, to access the financial and career opportunities of tech. Because even with all the challenges –and there have been many – the tech industry has been a fantastic career for me. The least I can do is open the door and send the elevator back down, so other people can access all that this industry and mindset enables you to do.”

One of the major challenges Stefflbauer has identified is algorithmic bias in recruitment.

“I think most of Europe is shockingly bad, but in Germany specifically, and the tech industry within that, it’s the use of photographs on CVs and the egregious lack of trained human resources that you find in a lot of start-up companies.

“It’s the rise of ‘people teams’, who often just outsource familiar, biased selection processes to the public. That makes Germany a pretty daunting job market, in terms of fairness and transparency in recruitment.”

Digital tools, mostly from the US, are embedding bias in recruitment processes, Stefflbauer argues, to the point where algorithms are down-ranking candidates based on keywords without any human intervention.

Does she think that is especially prevalent in the tech industry?

“It’s a pretty big problem because the pace of hiring has outstripped the capacity of most screening teams,” she says. “For efficiency’s sake, many companies use job boards that rank candidates to save time. The problem with algorithm-driven job boards is that the optimisation and efficiency is coming at the expense of candidates who can’t game the system.

“Academic articles have described the bias in these candidate-ranking algorithms. But what frightens me most is that when I talk to headhunters and recruiters, they don’t know how these algorithmic rankings work.”

Stefflbauer uses a concrete example: “I’m a chief financial officer [CFO] with my picture on my profile on LinkedIn, and I apply for a job as a CFO in Europe – but no-one knows how my profile gets ranked as ‘top10 per cent’ or ‘top 60 per cent’ of applicants.

“No-one knows whether, for example, a woman with gaps in her employment record will be ranked unfavourably next to a man with no gaps. No-one knows if my photograph, which a simple computer vision algorithm can use to identify me as female, and black, will make me an ‘outlier’ when compared with other CFOs.”

Data Capital: What excites the experts most about the potential for data in Scot...

For Stefflbauer, “blind” recruitment is still very much the exception: “There’s a very, very, very small minority, mostly in the UK, that are piloting anonymisation tools – stripping off information on age, photograph, gender, academic background, to see what we get with a completely fair and objective approach.

“At the moment, recruiters are just assuming, as is usually the case with algorithms, that recommendations are coming like magic from computers that have a lot of data. If it says these are the best candidates, they are the best candidates.”

What does Stefflbauer think FrauenLoop has achieved to tackle some of the systemic challenges she has identified? “The most profound thing is getting major players to acknowledge a problem in recruitment, and promoting tech to a broader segment of society. The other thing is improving understanding of how algorithms and bias work in recruitment in the digital age and what can be done to mitigate these biases.

“It’s about constantly stressing there is no typical profile for technology success and ensuring people who consume and depend on this technology get into the development and testing conversation.”

Stefflbauer says we are in a race to ensure tech and data can reach its full potential and be deployed positively: “At the moment, we are using automated systems to surveil and classify people. That’s what it comes down to – categorising and classifying humans.

“We’re acting as if this is inevitable, that the future is going to be automated and computers are eventually going to make all the decisions, with humans out of the loop. That’s simply untrue.”

She sees an opportunity for Europe to define a third way – “neither a laissez faire world playing on people’s credulousness… nor a hellscape where governments subject you to classification and ranking systems each time you purchase anything or leave your home”.

This third way, she says, involves regulation – building on GDPR, with support from the tech industry: “It’s about guardrails, working within sensible regulation – becoming accepted by the industry and those who use tech, and legally backed by the European Union.

“The EU’s AI [artificial intelligence] Act is due to come into force in 2024, and I hope this can create some kind of framework around which we aren’t categorised and deprived of individuality and choice when it comes to applying for jobs.”

Stefflbauer believes change must come soon, or faceless technology will shape the future, unchallenged.

“I think we’re at the turning-point where we’re almost facing irreversible data determinism. Working within sensible, human-rights-centred regulations is the only target that will ensure a fast enough solution to this dystopian future we’re otherwise looking at.”

Data Capital: Addressing universal inequalities