The Professionals series highlights outstanding experts across the PayPay Group. Behind PayPay’s rapid pace of feature releases is a group of professionals who bridge data and business, translating technology into real-world impact. This installment features Danyi Qian, a Data Scientist working in credit strategy, responsible for both risk control and growth across PayPay’s financial products. How did she move beyond analysis to become a direct driver of business decisions? We explore the reality of her career and what it means to practice data science at scale.
Danyi Qian
Financial Strategy Division, Financial Business Group
After beginning her career developing information systems for financial institutions at an IT vendor, She expanded her perspective as a Data Scientist at an IT-focused trading company, where she supported AI adoption and provided consulting at the intersection of technology and business. She joined PayPay in September 2023. Today, she works on credit risk and marketing model development for financial products including PayPay Shikin Chotatsu（PayPay Funding）, while also contributing to implementation, operations, and data integration at the very core of PayPay’s services.
From Advisor to Decision-Maker
What led you to join PayPay, and what are your aspirations today as a Data Scientist?
“Is my expertise truly moving the business forward?” Many engineers and data scientists ask themselves this question at some point in their careers—and I was no exception.
My journey began at university in Japan, where I studied economics and became fascinated by statistics as a way to understand societal dynamics through numbers. After graduating, I worked on financial system development at an IT vendor, building a strong foundation in the financial domain. I later worked as an AI consultant, helping large enterprises apply data and analytics across a wide range of industries. While working in consulting, I repeatedly encountered situations where even highly sophisticated models failed to reach real-world implementation due to organizational or decision-making constraints. Those experiences led me to want greater ownership—to be responsible for data from within the business.
That was when I encountered PayPay. During the interview process, I sensed a strong desire for co-creation from the business team—people who genuinely wanted to surprise the market and were looking for partners to run alongside them. I wanted to step away from the relative safety of making proposals and instead share responsibility for both success and failure. At PayPay, I was convinced I could experience both the rigor of financial systems and the speed of a large-scale tech service at a very high level. That conviction ultimately pushed me to join.
Driving Credit Decisions with Data
What is your current role and mission?
I work within the Financial Strategy Division, focusing primarily on consumer-facing credit risk and underwriting, while also being deeply involved in merchant-facing credit products and risk models. In short, my mission is to transform PayPay’s vast data assets into actionable credit risk decisions—balancing safety and speed while directly driving business growth.
A key focus of my work is PayPay Shikin Chotatsu (PayPay Funding – service details available in Japanese only), one of PayPay’s most advanced machine-learning-driven financial products. Here, the role of a Data Scientist goes far beyond building accurate prediction models. Credit is a domain where risk control and growth optimization are always in tension. On one hand, it demands strict discipline—mistakes are not tolerated. On the other, PayPay’s business requires speed and scalability. Designing where to draw that line, and doing so based on data rather than intuition, is one of the most challenging and rewarding aspects of my role.
When I first joined, my primary responsibility was model development. Today, my role has expanded upstream: consolidating data during the planning phase, defining requirements, and aligning priorities across teams. I sit at the same table as product managers and business leaders, discussing timelines, expected impact, and trade-offs—and then seeing initiatives through to implementation. That sense of ownership is central to how I approach my work.
Evolving PayPay Shikin Chotatsu Through Data
Which recent initiatives left the strongest impression on you?
Without question, the launch and continuous improvement of PayPay Shikin Chotatsu. While the core design was already in place when I joined, I was involved in everything from improving model accuracy to defining infrastructure requirements for live operations. Because this was PayPay’s first financial service to fully leverage machine learning, its success quickly built momentum. Business teams began approaching us with requests like, “Can we model this next?” or “Can we optimize decision thresholds using this metric?” Seeing my models directly reflected in merchant offer conditions—and ultimately in revenue numbers—was something I had rarely experienced as a consultant. It gave me a tangible sense of what it means to practice data science inside a business, where models directly shape real-world outcomes.
Another pivotal experience was a joint project with PayPay Bank. Using bank-specific data, I was involved from the upstream stages—deciding which models to build and which metrics should guide decision-making. Defining shared KPIs across different organizational cultures and visualizing them through dashboards felt like connecting organizations through data.
Balancing Speed and Discipline
What principles do you prioritize when driving projects forward?
Constructive dialogue across departments with different perspectives is critical. In one project, our team wanted to implement immediately, while the product team faced resource constraints. Instead of settling for compromise, we stepped in to support part of the implementation ourselves—driven by strong conviction that the initiative was essential for the business. The result was a highly successful experiment.
At the same time, Data Scientists must also know when to apply the brakes. In credit, revenue and risk are always in a trade-off relationship. Especially when business momentum is high, it becomes even more important to rely on data—not intuition—to assess the situation.
Before adopting or modifying machine learning models, we validate them in a way that does not affect existing operations, estimating expected performance and risk in advance. Rather than asserting “We should use this model,” we align first on facts: “Under these conditions, this impact is expected.” We also standardize evaluation criteria and share results visually. By using the same metrics and reports as the risk management team, discussions shift away from subjective judgment toward shared, data-driven decision-making. What matters most is not stopping at risk identification. By proposing data-backed alternatives—defining where it’s safe to challenge and what level of risk is acceptable—we enable the organization to move forward with confidence.
Diversity of Expertise That Fuels Challenge
How would you describe your team culture?
In one word: diversity. Our team includes members with deep hands-on credit experience at overseas banks, alumni of internet service companies, seasoned engineers from IT vendors, and Data Scientists seconded from the data strategy group.
Meetings often blend Japanese, English, and Chinese. Being trilingual myself, I frequently bridge communication between engineers and the Japanese business team. When someone tells me, “Thanks to you, the technical intent finally came through,” I’m reminded that my background directly contributes to the team’s effectiveness. We also value open knowledge-sharing and actively avoid siloed expertise. Pair work is common, and weekly knowledge-sharing sessions cover everything from advanced Snowflake usage to applications of location data.
Many team members are parents. The environment is highly flexible around life events—when something comes up, the team supports one another. Even in a hybrid setup, there’s no sense of isolation; it’s a place where professionals genuinely rely on each other.
Beyond Models: Becoming a Business-Driving Data Scientist
What makes working as a Data Scientist at PayPay compelling?
What stands out most is the ability to experience data science not as theory, but as an integral part of the product lifecycle. The scale of data is extraordinary—real-time data generated by 72 million users (as of December 2025) and millions of merchants. That scale alone presents a significant challenge, and at the same time, many areas remain untapped.
What’s particularly interesting is that PayPay’s data allows us to see users not as isolated points, but as multi-dimensional entities. By combining payment history, attributes, and behavioral patterns, we gain insights that go far beyond balances or credit scores. This makes it possible to clearly define which value should be delivered to which user segments—and to have concrete, data-grounded discussions around those decisions.
Another strength is working alongside Machine Learning Engineers (MLEs). Data Scientists focus on modeling and problem framing, while MLEs design pipelines and operational infrastructure. By overlapping our domains, we gain exposure to both sides, making it an exciting environment for those with strong technical curiosity.
PayPay is also quick to adopt new AI tools. Recently, development support tools like Cursor have been rolled out widely, significantly improving productivity. This allows us to spend less time on pure implementation and more time on defining objectives, prioritization, and understanding users—further elevating the role of the Data Scientist.
Turning Change into Momentum
What are your future goals and vision?
The credit domain at PayPay still has tremendous growth potential. My challenge is to fuse the operational stability expected of financial institutions with PayPay’s culture of bold experimentation.
Rather than pursuing academic research alone, I want to sharpen my business sense and become a Data Scientist who can influence decisions across domains. By deepening expertise in credit and marketing, I aim to move beyond analysis and step into a role that leads projects as a true business partner. It’s not about obsessing over a 1% improvement in model accuracy—it’s about understanding how that 1% changes user experience and moves revenue. That’s the level of responsibility I aspire to carry.
What Moves Careers Forward Is the Will to Take on Challenge
What qualities do successful people at PayPay share?
First and foremost, the ability to embrace change. At PayPay, decisions made in the morning may change by evening. Enjoying that dynamism and adapting flexibly is essential. Just as important is the ability to communicate effectively. Our work doesn’t end with building models—it’s about translating insights into product and business decisions. Collaborating with product, marketing, and risk teams to derive better options requires open dialogue. Those who collaborate beyond their own domain tend to leave the greatest impact.
From a technical standpoint, Python and SQL proficiency and the ability to explain model logic are expected. But more than that, we value people with a clear sense of direction—those who want to wrestle with massive datasets you can’t find anywhere else and bring them into real-world impact.
When I first came to Japan, curiosity outweighed anxiety. My experience at PayPay feels much the same. How far can I push the boundaries of my expertise? How can this vast data create new value? If you want to apply data within a business—and move society with your skills—there are few places more stimulating than PayPay. Let’s build the future of finance together.
