How Humans in the Loop is Building the Workforce of the Future in Conflict-Affected Regions

Iva Gumnishka of Humans in the Loop tells us about the human workforce for continuously better computer vision.

First of all, how are you and your family doing in these COVID-19 times? 

Iva Gumnishka: My family and I have been fortunate to have not fallen sick yet or to have had a very mild version of the virus. The pandemic in Bulgaria where our HQ is has not had effects as dire as in our countries, and the lockdown has also imposed more relaxed measures compared to the stringent rules elsewhere.

Tell us about you, your career, how you founded Humans in the Loop.

Iva Gumnishka: Humans in the Loop is a social enterprise founded in 2017 with the mission to channel remote online work opportunities to refugees and conflict-affected people. We are specialized in data annotation for the computer vision industry, which means we get to collaborate with and support many exciting projects in the field of autonomous driving and flights, medical AI, agriTech, smart industry, robotic vision, and others. Essentially, we perform the manual labeling and verification these companies need in order to train their ML models. Since 2017, our company has provided employment to more than 400 people across Bulgaria, Turkey, Syria and Iraq.

After graduating with a degree in Human rights from Columbia University, I was driven by the desire to go back to my home country Bulgaria and bring more innovation to the sphere of refugee integration and livelihoods by connecting vulnerable people with the jobs of the future. We found a great niche which has two advantages: it has a very low barrier to entry, and it’s very appropriate for lower-skilled people who don’t have other work opportunities; at the same time, it’s a very dynamic industry in which our workers acquire valuable skills and knowledge about AI systems and how to train and audit them.

How does Humans in the Loop innovate? 

Iva Gumnishka: As a dataset annotation and collection company, we serve our clients in order to deliver high-quality gold standard data that they can use to train their computer vision models. However, we also see the need for continuous human supervision and monitoring of models currently being deployed. This would benefit the companies greatly by mitigating model drift and ensuring models are up-to-date and bias-free.

We are focused on educating our clients and raising awareness about the risks of bias in AI and how it can be eliminated through better data collection and annotation, and services such as model auditing and error analysis. Hopefully, this will lead to the creation of AI systems that are more ethical and built with diversity and inclusion in mind.

How does the coronavirus pandemic affect your business finances?

Iva Gumnishka: We are grateful that the pandemic has not affected our business and in fact, there is even a bigger need for our services. On the one hand, there are more and more people in the communities we serve have found themselves in a precarious situation after losing their jobs and facing the need for upskilling in order to access the online job market. On the other hand, more and more companies are developing AI solutions that respond to the pandemic’s new challenges: mask detectors for ensuring public health, AI solutions for detecting coronavirus on radiology imagery, etc.

Did you have to make difficult choices regarding human resources, and what are the lessons learned?

Iva Gumnishka: At the beginning of 2019 we switched from having an in-house full-time team to working with annotators remotely on a freelance basis, which seemed to be a much more attractive proposition to many of them which wanted to have flexible work hours in order to combine their work with other responsibilities. Little did we know that in 2020 this would end up being a great benefit for us because we adjusted extremely fast to the new conditions and all of our workers were already equipped with the tools, technology and soft skills in order to perform their work remotely. This was a great benefit for our clients as well since there were no interruptions in our services.

How did your customer relationship management evolve? Do you use any specific tools to be efficient?

Iva Gumnishka: Our customer relationship management has not changed during the pandemic.

Did you benefit from any government grants and did that help keep your business afloat?

Iva Gumnishka: Governmental grants were only available to businesses who had lost a considerable size of their revenue during the spring months of 2020. Fortunately, this was not the case for us, and our business continued to grow despite the pandemic.

Your final thoughts?

Iva Gumnishka: As we are already adapting to the “new normal”, we at Humans in the Loop are seeing a growing need for integrating communities with lower digital skills into the online remote work market, so that they are able to enjoy the safety of working from their homes and the advantages of the expanded access to global opportunities. We are thinking of ways to upskill people to access freelance marketplaces and provide data curation work, which is becoming a new type of profession. We envision that this sort of professional “humans-in-the-loop” will in the future be specialized to train and teach machines how to think like a human, and we are excited to create these types of jobs in conflict-affected regions and countries.

Your website?

https://humansintheloop.org