Fintech Interviews Machine Learning

Global Fintech Interview with Peter Barcak, CEO and Co-Founder at credolab

Global Fintech Interview with Peter Barcak, CEO and Co-Founder at Credolab

Hi Peter, welcome to our Fintech Interview Series. Please tell us about your fintech journey so far.

I’ve been focused on risk management and leading teams in banking for over 20 years. Having started at Citibank, then I moved to Intesa Sanpaolo Europe, where I created retail risk management departments from scratch. Later I joined my first startup, Platinum Bank in Ukraine, as a Chief Risk Officer, where I experimented with alternative data sources such as telco or social media data to make informed credit risk decisions. This led me to co-found credolab in 2016 with the goal of providing a data-driven solution that promotes financial inclusion by helping lenders better understand their customers and reduce loan rejection rates.

Today, our platform is in high demand across various verticals, including risk management, fraud prevention, marketing segmentation, and more.

With credolab embedded technology, our clients can gain rich insights and scores that lead to clear and measurable benefits. We have already worked with over 200 corporate clients across 9 verticals, and despite the global downturn, 2022 was a year of tremendous momentum for us.

Credolab is at the forefront of developing predictive ML-driven analytics to enable faster and better credit scoring, marketing targeting, or fraud detection. Could you highlight its technical applications?

Credolab’s platform leverages more than 70,000 anonymized data points to detect behavioral patterns, such as battery usage, the most frequently downloaded app category, the total time spent applying, speed of the scrolling or the number of times “copy” and “paste” was used on questions the user should know the answer to.

Leveraging the experience of over 180 million digital footprints by scoring more than 20 million people to date, credolab’s proprietary machine learning algorithms help identify pockets of opportunities within the same user base that was thus far ignored. With the right tech tools and expertise, we can help mine deep behavioral insights of mobile apps and website users, ultimately improving financial services for people around the world. As for how it all works, without being too technical, our technology requires being embedded in the front end of clients’ mobile apps or websites and connecting with our API. The main effort required by our clients is simply the prioritization of the integration. The “IT heavy lifting” is done by credolab. 

Also Read: Global Fintech Interview with Sagar Rajgopal, President and Chief Customer Officer at Ubiquity

Why are AI models still a “black box” for financial institutions, and what steps can be taken?

Despite all advancements, the lack of explainability in AI models is still a concern in traditional banking and insurance with their conservative mindset. There have been several high-profile examples of bias in AI models that have raised concerns with consumers and regulators. This hesitation stems from worries about the explainability of the results, particularly if it may bring bias to the general model in terms of ethnicity, gender, or even ZIP code. I am strongly convinced that the issue of AI explainability can be dealt with by starting with clear principle “garbage in – garbage out”.  It applies to the data used for “feature construction” and data used for “target calculation”.

If you create useless features or variables with data leaks (the use of information from outside the training dataset that would not normally be available at the time of prediction) you are likely to overestimate the predictive ability of your model in production mode.

To get stable and reliable results, it is crucial to define the target consistently for all approved loans, ensuring that each loan has an equal chance of becoming good or bad. If the basics are done correctly, the choice of modelling algorithm becomes irrelevant. Ultimately, the apparent crisis of AI explainability vanishes if the features are constructed correctly and properly calculated target variables are used.

How can AI-based solutions help improve credit scoring thereby benefiting both the borrowers as well as the lenders?

When we started working with financial institutions as clients, it turned out that most lenders rely on out-of-date data offered by credit reporting agencies (CRAs) or bureaus as part of the creditworthiness assessment of borrowers. However, the bureaus can only provide scores for applicants with an existing credit history, usually those in the middle to upper-income groups. Without that same data, many people have remained financially excluded.

With credolab technology, we help lenders reach more people and assess not past but predict future behavior. We also work with credit bureaus as clients, so we see a desire for inclusivity and improved, more accurate scoring throughout the industry. 

Also Read: Global Fintech Interview with Al Morris, Chief Executive Officer at Koii

As per the recent statistics, there is a hacker attack every 39 seconds, what are your strategies to deal with this situation where security is being compromised?

In credolab, we have a multi-layered security approach that includes network security, application security, and physical security measures to protect our systems and data. We regularly perform vulnerability assessments and penetration testing to identify potential weaknesses and address them promptly. We also have dedicated teams for monitoring and analyzing security-related events and anomalies to quickly detect and respond to potential threats. Furthermore, we follow industry best practices and comply with relevant regulations and standards, such as the General Data Protection Regulation (GDPR) or ISO 27001, to ensure the security and privacy of our client’s data.

Could you please explain how you are helping lenders of all kinds to make better credit decisions leveraging AI models?

Unlike Meta or Google, we don’t collect data to show targeted advertising. We give people a fairer assessment for getting credit or other financial services. Credolab uses its proprietary embedded technology and a scoring engine based on machine learning algorithms to analyze digital footprints and generate highly predictive intent scores. Lenders can use this information to make better credit decisions. We ensure full compliance with privacy regulations by focusing on privacy-consented metadata, which is completely anonymized and highly secured. We never share raw data with clients, and all metadata is stored in JSON format on our cloud. 

For example, the collaboration between credolab and one of our partners, Provenir, has already contributed to improving financial inclusion in LATAM, where only 42% of adults in the region say they have access to financial services in 2021.

What is one piece of advice that you would like to give to our readers for their financial security?

Keep track of your transactions, regularly review your statements, and immediately report any suspicious activity to your financial institution. Additionally, regularly checking your credit report can help you spot any errors or fraudulent activity and allow you to take steps to correct them. By being vigilant and proactive in monitoring your finances, you can help prevent and minimize the impact of financial fraud and protect your financial security.

Also Read: Global Fintech Interview with Timothy Rooney, President at Marygold & Co

What are your predictions for the fintech industry for 2023-2024?

One of the major trends that will continue to shape the fintech industry in the coming years is, of course, the increasing use of AI and machine learning. We will likely see a surge in the growth of niche and specialized AI solutions for all financial sectors.

For example, Bloomberg already released a research paper detailing the development of a new large-scale generative AI model that has been specifically trained on a wide range of financial data to support a diverse set of natural language processing (NLP) tasks within the financial industry.

I have read various predictions about how many jobs could be cut by implementing a variety of AI solutions. As any member of the human race, I am, of course, a bit frightened by such prospects. But as with all innovations and breakthroughs, from the invention of the printing press to man moving from horse to the car, it certainly only moves our civilization forward and outweighs the drawbacks.

Furthermore, I expect fintech companies to focus more on financial inclusion, working to provide access to financial services for underserved populations worldwide. This will require a concerted effort from both the private and public sectors to overcome the challenges of reaching underserved communities and ensuring that financial services are accessible and affordable.

Overall, the fintech industry will continue to evolve rapidly, with innovations and developments emerging regularly. The key to success in this dynamic landscape will be to stay agile and adaptable, continually pushing the boundaries of what is possible with technology to create more efficient and accessible financial services for all.

Thank you, Peter! That was fun and we hope to see you back on soon.

[To share your insights with us, please write to]

After over 20 years in senior roles at multi-national banks and startups, Peter Barcak is an experienced leader who has had a tangible impact on industries dealing with risk calculation. Before launching credolab in 2016, Peter held senior roles in leading banks – including BNP, Citibank, Intesa Sanpaolo Group, and Platinum Bank. Peter is a high-integrity, inspiring, and motivational leader with a real passion for developing people. He is known for his ability to envision and produce successful outcomes in complex situations, and he is driven by the belief that our financial systems can be more inclusive and profitable. 

сredolab Logo

Founded in 2016, credolab is today’s largest developer of bank-grade digital scorecards and data enrichment solutions. The company provides lenders, risk officers, and marketers with a previously untapped, highly-predictive source of behavioral data: privacy-consented and anonymous smartphone metadata. Leveraging the experience of over 180M digital footprints by scoring 20M+ people to date, the company’s toolkit is already in use for over 200 corporate clients across 9 verticals, including banks and neobanks, BNPL, digital lenders, credit bureaus, ride-hailing and EWA apps, insurance companies, or crypto lenders. Credolab is headquartered in Singapore with offices in London and Miami.

Related posts

Featurespace Launches Automated Deep Behavioral Networks

Fintech News Desk

Zions Bancorporation Joins Real Time Payments Network

Fintech News Desk

New Study Reveals the Struggles that Many Businesses Face in Getting Paid During the COVID-19 Pandemic

Fintech News Desk