Qfin Holdings, Inc. is an AI-empowered Credit-Tech platform in China.
Leveraging its sophisticated machine learning models and data analytics capabilities, the company provides a comprehensive suite of technology services to assist financial institutions, and consumers, and SMEs in the loan lifecycle, ranging from borrower acquisition, preliminary credit assessment, fund matching, and post-facilitation services, with the Qifu Jietiao app (previously known as the 360 Jietiao app) as the company'...
Qfin Holdings, Inc. is an AI-empowered Credit-Tech platform in China.
Leveraging its sophisticated machine learning models and data analytics capabilities, the company provides a comprehensive suite of technology services to assist financial institutions, and consumers, and SMEs in the loan lifecycle, ranging from borrower acquisition, preliminary credit assessment, fund matching, and post-facilitation services, with the Qifu Jietiao app (previously known as the 360 Jietiao app) as the company's primary user interface. The company is dedicated to making credit services more accessible and personalized to consumers and SMEs through Credit-Tech services to financial institutions, whereby the company deploys its technology solutions to help financial institutions identify the diversified needs of consumers and SMEs, effectively access prospective borrowers that are creditworthy through multi-channels, enhance credit assessment on prospective borrowers, and manage credit risks and improve collection strategies and efficiency, among others. With user insights distilled from long-term engagement with users across life and business scenarios enabled by AI and data analytics, the company's technology solutions empower financial institutions across different stages of the loan lifecycle, enabling them to extend the reach of services and satisfy the financing needs of consumers and SMEs, and deliver to users more accessible credit services. In turn, the company primarily derives service fees from its technology solutions to financial institutions. The company is building a comprehensive Credit-Tech service platform that encompasses an extended user lifecycle and promotes financial inclusion. As of December 31, 2024, the company had 56.9 million users with approved credit lines, accumulatively.
Services
The company matches underserved and unserved users with credit demand to a diversified pool of financial institutions with credit to supply, through both credit-driven services and platform services.
Credit-driven Services
Under the credit-driven services category, the company matches prospective borrowers with financial institutions and empower financial institutions in borrower acquisition, credit assessment, fund matching and post-facilitation services. Loan products offered under this line of services are primarily funded by the company's financial institution partners, with the remainder extended by Fuzhou Microcredit, which is licensed to conduct micro-lending business in China or trusts and ABSs. In both cases, the company bears credit risks of the loans. For loans extended by the company's financial institution partners, it provides guarantees against potential defaults. Such contractual guarantee arrangement is underwritten either by the licensed VIEs, or third-party licensed guarantee companies or insurance companies, to which the company may provide back-to-back guarantee at their request. With respect to loan facilitation services for loans funded by financial institution partners, the company charges service fees directly from its financial institutional partners pursuant to pre-negotiated terms based on the contractual agreements that vary from case to case. The company's service fee rate is typically the difference between the loan pricing rate, which is set by the financial institutions, and a fixed rate negotiated between the company and the respective financial institutions. For loans funded by Fuzhou Microcredit, it charges borrowers interest fees, which reflects a number of factors, including the credit profile of the borrowers, the availability of funding and the associated funding cost, and the tenor of loan products, among others.
Platform Services
The company's platform services include loan facilitation and post-facilitation services through its capital-light model, intelligent marketing services to financial institution partners under Intelligence Credit Engine, referral services and other technology solutions. The company does not take credit risk under platform services. For the year ended December 31, 2024, loans facilitated under the company’s platform services accounted for approximately 53.0% of its total loan facilitation volume.
Capital-light Model
The company launched its capital-light model in 2018 with the focus on implementing its strategic transition from a traditional risk bearing loan facilitator to a technology enabler. Under its capital-light model, the company facilitates transactions between prospective borrowers and its financial institution partners through a suite of technology-enabled services spanning across the loan lifecycle, including borrower acquisition, technology empowerment in credit assessment, and post-facilitation services, such as loan performance monitoring and loan collection. Under its capital-light model, the company historically provided limited guarantee to certain collaborating insurance companies in the event of bankruptcy and certain financial institution partners pursuant to their internal requirements. For loans facilitated under its capital-light model, the company generates income through service fees charged to financial institution partners according to pre-negotiated terms that vary from case to case. The company's service fee rate is typically a certain percentage of the pricing rate that is set by the financial institution partners on the loans to borrowers.
Intelligence Credit Engine (ICE)
ICE is an open platform that offers financial institution partners and other lending platforms intelligent marketing services. For loans facilitated through ICE, the company matches prospective borrowers with financial institution partners based on comprehensive data analysis and cloud computing technologies, and assists financial institution partners with preliminary credit screening of borrowers, but do not provide advanced credit assessment. The company earns pre-negotiated service fees from financial institution partners and does not bear credit risks. The company's service fee rate is typically a certain percentage of the pricing rate that is set by the financial institution partners on the loans to borrowers, and the service fee rate is subject to negotiations with the financial institution partners and varies from case to case.
Referral Services
As different financial institution partners prescribe different metrics assigned with various values in granting credit line approvals to prospective borrowers, some users fail to match the criteria of the company's financial institution partners and are rejected by them. However, such borrowers may still be within the target borrower group of other online lending companies. To offer more products and better user experience to its users and maximize the value of user traffic on its platform, the company provides referral services primarily to other online lending companies in line with industry practice and earns referral fees. The company considers referral services to be supplemental in nature to its loan facilitation services.
Other Technology Solutions
In 2020, the company began to offer financial institutions on-premises deployed, basic risk management SaaS solutions, or RM SaaS solutions. Since 2023, the company started to offer end-to-end technology solutions, including user acquisition, risk management, credit operation, and post-facilitation services to financial institutions based on on-premises deployment, SaaS or hybrid model, which it refers to as total technology solutions. Basic RM SaaS solutions and total technology solutions together are previously referred to as other technology solutions. Integrated with its credit assessment insights and algorithms, as well as other proprietary technologies, other technology solutions help financial institution partners acquire borrowers and improve credit assessment results. Under this model, the company typically takes technology service fees or consulting fees for the corresponding technology solutions elected by the financial institutions.
In terms of accounting treatments, under credit-driven services, the company either provides guarantees for loans funded by financial institution partners, which are recorded as off-balance sheet loans, or funds loans through trusts and ABSs or Fuzhou Microcredit, which are record as on-balance sheet loans. Under platform services, all loans facilitated through the company's platform are recorded as off-balance sheet loans. The company has a large balance of guarantee liabilities during the year ended December 31, 2024, as it provides guarantees under credit-driven services. The company also has a large balance of accounts receivable and contract assets, as well as financial assets receivable during the same period, mainly arising from off-balance sheet loans, as well as loans receivable, mainly arising from on-balance sheet loans. The company has established an evaluation process designed to determine the adequacy of its impairment allowances and guarantee liabilities, and an allowance for uncollectible receivables and contract assets based on estimates that incorporate historical delinquency rates by vintage and other factors surrounding the credit risk of specific underlying loan portfolios.
Products Offered to Users
The company's core product offered to users is an affordable, digital revolving line of credit allowing multiple loan drawdowns, with a convenient application process and flexible loan tenors. The company's products are provided under the Qifu Jietiao brand.
The company's engagement with prospective consumer borrowers begins with a credit application, which typically takes a few minutes. Once approved by its financial institution partners, a prospective borrower is granted a line of credit, typically with a principal amount ranging from RMB1,000 to RMB200,000, for drawdowns based on specific needs, with an amount typically between RMB500 and RMB200,000. When an approved borrower makes a drawdown request, the company performs a preliminary credit assessment on such borrower to ensure his or her continued qualification for drawdown before the request is transmitted to its financial institution partners for their independent final risk assessment and loan disbursement approval. Once a drawdown is approved, a borrower may elect a loan tenor best suited for his or her financial needs, in fixed terms of one month, three months, six months, twelve months, eighteen months, twenty-four months, or thirty-six months, to be repaid in monthly installments. The average amount of approved credit line for each borrower in 2024 was RMB14,236. In the instance where the company provides guarantee services, the guarantee services are provided throughout the loan tenor. The company is also offering other payment terms, such as repayment at any time with a fixed daily interest.
Service Process and Operation Flow
With the focus on empowering financial institution partners and serving consumers and SMEs, the company's platform offers services covering the entire loan lifecycle. Credit-driven services and the company's capital-light model follow the same service process and operational flow from credit line approval to loan drawdown, and differ only in the post-facilitation stage, where under credit-driven services in which it bears credit risks, it makes guarantee repayments to its financial institution partners if needed. For ICE, as the company provides financial institution partners intelligent marketing services, it mainly conducts preliminary credit screening of prospective borrowers during the credit line approval stage, therefore participating in fewer steps in the loan lifecycle than it does under credit-driven services and its capital-light model.
Stage 1: Credit line approval
Step 1: Paperless Credit Application: For new users, the company's service journey begins with such users' registration of an account on its platform by providing it with certain basic information and authorization to collect other information for fraud detection and credit assessment, among others. The credit application process typically takes a few minutes, after which the company initiates a user portrait profiling, fraud detection and credit assessment process.
Step 2: Portrait Profiling, Fraud Detection and Credit Assessment: The company deploys the Argus Engine to build a prospective borrower profile for fraud detection and credit assessment. Drawing on the company's database, AI-enabled credit assessment system, Argus Engine, and understanding through interactions with a broad user base, it is able to develop a more accurate and comprehensive prospective borrower portrait. Once an applicant passes the fraud detection test, the company initiates a comprehensive credit assessment and generate a proprietary credit score for the applicant under credit-driven services and its capital-light model, or conduct only preliminary credit screening under ICE.
Step 3: Recommendation and Matching: Through its workflow system CloudBank, under both credit-driven services and its capital-light model, the company then recommend the prospective borrower's profile along with pricing recommendation to its financial institution partners and share the results of its preliminary credit assessment with them to facilitate their final risk management and credit decision making including loan tenor, approved credit line, and other key terms of a loan product. For ICE, the company only recommend prospective borrowers to financial institution partners based on the results of preliminary credit screening, and do not provide pricing recommendations.
Step 4: Final risk management and credit decision by financial institutions: The financial institution partners conduct final risk management and make their credit decisions based on their respective credit process and regulatory guidelines.
Step 5: Notice on credit line approval. Following their final risk management, each financial institution partner will respond to the company’s workflow system indicating approval or rejection, and in the case of approval, their maximum level of credit exposure. Upon receiving the credit approval decision from financial institution partners, the company passes such information to prospective borrowers through its platform.
For ICE, as the company only recommends prospective borrowers to financial institutions after preliminary credit screening.
Stage 2: Loan Drawdown
Once a credit line is granted, a prospective borrower may request a drawdown at any time, subject to the credit limit approved by the financial institution partner. Upon receipt of a drawdown request, the Argus Engine conducts a streamlined credit assessment to ensure the prospective borrower's continued qualification for drawdown and notifies the company's financial institution partners of the drawdown request, which complete their final risk management and reach a drawdown decision. The company undertakes to notify the borrower the drawdown decision and the financial institution partner that is matched with the borrower will disburse loan to the borrower. Once the principal of the loan is transferred to the borrower, the company recognizes revenue from loan facilitation services for services provided to the financial institution partner.
For ICE, although the prospective borrower's drawdown application is made through its platform, the application is directly sent to its financial institution partner through the application programming interface (API) without it processing of the information in any way.
Stage 3: Post-facilitation services: continual credit profile monitoring and collection
Robust data analytics technologies have enabled the company to continuously monitor the credit profiles of borrowers. After a borrower makes a loan drawdown, the company's Argus Engine tracks his or her borrowing and repayment activities, and automatically adjusts such borrower's credit profile on an ongoing basis. Borrowers typically make repayments to the company's financial institution partners through third-party payment platforms rather than through its platform. The company recognizes revenues from post-facilitation services on a straight-line basis over the term of the underlying loans. The company typically collects pre-negotiated service fees (inclusive of fees for loan facilitation services, post-facilitation services and guarantee service fees, if applicable) from financial institution partners on a monthly basis as borrowers make repayments over the term of the underlying loans.
For loans facilitated under ICE, the company also provided limited collection services to a small portion of financial institution partners based on their special requests.
Credit Demand
Target User
In consumer Credit-Tech market, the company targets the large and growing Chinese population of users who typically has stable income with promising growth potentials and has greater user lifetime values but are underserved or unserved by the traditional financial institutions. Prospective borrowers are generally drawn to its platform for supplemental credit solutions.
In the SME Credit-Tech market, the company’s products mainly aim to serve SMEs with an annual operating revenue below RMB5 million, which are typically granted with credit line below RMB1 million.
The company has established a large base of loyal creditworthy users. As of December 31, 2024, the company had 56.9 million cumulative users with approved credit lines in the aggregate, among which 60.3% had credit cards, mortgage loans or auto loans and 43.2% were between 25 to 35 years old. The company's repeat borrower contribution was 93.1% for the year ended December 31, 2024.
User Acquisition
The company strives to diversify the network for user acquisition, which comprises online advertising on channels operated by leading internet companies, embedded finance cooperation with online platforms with heavy user traffic, offline promotions and referral programs with other platforms and 360 Group.
Online Advertising
The company partners with leading internet traffic platforms to acquire borrowers via online advertising. The company is improving its targeted marketing capabilities by leveraging data analytics so that it can place advertisements to intended users who fit into its target borrower profile more effectively. The company has also developed analytics algorithms in collaboration with channel partners based on the anonymous user information aggregated from such channel partners so that users of the channel partners with credit needs can be directed to its platform with improved precision and efficiency. The company intends to continue optimizing its proprietary AI and data analytics systems and expand the network of channel partners to improve user acquisition efficiency.
Embedded Finance Model
In 2020, the company started cooperating with leading online platforms with heavy user traffic under the embedded finance model. These platform partners include, among others, leading short-form video platforms, e-commerce platforms, ride-hailing companies, and smartphone companies. In 2024, the company started to explore collaborations with financial institutions to engage their existing customer bases, leveraging their proprietary traffic alongside its differentiated pricing and service capabilities to expand the breadth and depth of its user coverage. Under this model, the company embeds its credit assessment, data analytics, and other proprietary technology solutions within the partnering internet platforms and financial institutions. Therefore, credit services used by end users of its partnering platforms will ultimately be provided by the company. Through embedded finance, the company is able to reach more users effectively while empowering its partnering platforms and financial institutions to improve user experience and further unleash the monetization value of their user base. The company has become the Credit-Tech service partner of many leading online platforms and financial institutions across various categories, gaining access to a large number of internet users across consumption scenarios for potential conversion into borrowers. As of December 31, 2024, the company had partnered with 56 leading online platforms and financial institutions cumulatively, and embedded finance has become an important user acquisition channel for it.
Offline promotion and borrower referral programs
In the meantime, the company conducts offline sales and marketing activities to promote its products and services in specific regions and for specific products. In addition, the company continues to acquire new users through borrower referral programs.
360 Group Channels
Historically, the company collaborated with 360 Group in several aspects of user acquisition. Benefiting from the collaboration, which enables its mobile app to be showcased on 360 Group's products' user interfaces, the company has been able to connect with 360 Group's user base. In recent years, however, prospective borrowers acquired from 360 Group has contributed significantly less to its business, as its user acquisition channels continue to diversify.
Credit Supply
The company has a stable and diversified base of funding partners. The company primarily relies on its financial institution partners, including national and regional banks and consumer finance companies, to fund its credit products. From time to time, the company also funds a certain percentage of loans through Fuzhou Microcredit. With sufficient and strong funding commitment from its financial institution partners, the company has the flexibility to recommend suitable products to borrowers with different combinations of funding sources depending on market conditions.
Financial Institutions
The company's financial institution partners are mainly national and regional commercial banks and consumer finance companies. The value the company adds to its financial institution partners includes efficient borrower acquisition through online and offline channels, credit assessment technology empowerment, post-facilitation services and risk-adjusted returns throughout economic cycles, among others. The company's technology infrastructure helps enhance financial institution partners' risk management, providing them with a more seamless and real-time risk management experience.
In certain special cases and as mutually agreed upon by the company and a small number of financial institution partners pursuant to their internal business requirements and procedures, some of the loans facilitated through its platform are funded by and disbursed indirectly through trusts, which also provide the company with more flexibility to utilize the funds from the trusts for loan facilitation within the specified time frame and are in line with the industry norms.
As of December 31, 2024, the company had established partner relationship with a total of 162 financial institutions cumulatively, including national and regional banks and consumer finance companies, across 26 provinces and autonomous regions of provincial level and 70 cities in China.
Fuzhou Microcredit
In March 2017, Fuzhou Microcredit was established, which has obtained the regulatory approval and micro-lending license to originate loans.
Alternative Funding Initiatives
The company has explored and expects to continue exploring alternative funding initiatives, which include standardized capital instruments such as the issuance of ABSs and ABNs.
Credit Assessment
The company's credit assessment technology solutions are built upon a comprehensive database, a sophisticated credit profiling engine, and an efficient post-facilitation service process. With its technology empowerment, financial institutions conduct core risk management and credit approval independently to achieve better risk management.
Comprehensive Database
Large volume of high-quality data is a key factor differentiating Credit-Tech platforms. With users' consent to its use of their data, the company has developed a comprehensive database comprising a large volume of reliable information, including among others, a user's credit history, credit lines granted by banks, consumption pattern and past repayment behavior, that are relevant to the assessment of a given user's credit risk against future borrowing. The company develop sits database and build user profile primarily with its first-hand and proprietary data. Meanwhile, the company also partners with third-party data providers to enrich its database of credit information. For example, the company has access to the People's Bank of China's credit reporting system, which allows the company to retrieve and submit data on borrowers' credit profiles.
Credit Assessment Engine
The success of its business relies on the effectiveness of its credit profiling systems. The ‘brain’ of its credit profiling systems is its Argus Engine. The company's Argus Engine integrates user database, AI-powered data analytics, and expert experience based on AI technologies, such as machine learning and deep learning, into comprehensive models. It allows the company to effectively recognize and infer the patterns and relationships between information nodes and develop user profiles more accurately without substantial human intervention. For example, its Argus Engine is capable of automatically and continually training its algorithms with data in real life, and iterating and refining the precision of its profiling and decision making across the lifecycle of a loan. In addition, the company has equipped the Argus Engine with a number of cutting-edge technologies in the area of AI, including machine learning and deep learning, which enable a more effective screening of fraudulent application and a more precise profile buildup. For another instance, the company has programed large-scale social network (knowledge graph) into its Argus Engine for fraud detection, which empowers the company to comprehensively map and reason about connections between its users, and therefore more effectively identify organizational fraudulent behaviors. Leveraging its three core functions of anti-fraud, credit assessment and risk alert, Argus Engine helps the company effectively builds user profile, conduct overall credit assessment for each prospective borrower and detect frauds, thereby lowering the possibility of loan delinquency.
Behavior Analysis and Fraud Detection
The Argus Engine is deployed to conduct fraud detection and initial credit screening of a prospective borrower, generating an F-Score which is a proprietary metric quantifying potential fraud risks of the borrower. Through its Argus Engine, the company seamlessly combines data aggregation with fraud detection capabilities as follows.
Identity Authentication: The company uses facial recognition technology and other tools and processes to verify the identity of a prospective borrower, denying those applications with what the company to be false identities.
Blacklist Filtering: The company maintains a real-time list of suspicious devices and accounts referred to as a blacklist and to which it has automated access. The company refers to the blacklist, as well as fraud records provided by third-party institutions to filter prospective borrowers with high fraud risks.
Telecommunication Fraud Prevention: The company's anti-telecommunication fraud system integrates black or gray list, AI powered source tracking technologies, as well as real time transaction and risk monitoring models. This system enables fraud prevention across the entire lending process, from pre-facilitation borrower acquisition to post-facilitation services. Its telecommunication fraud prevention mechanism features fraud risk alert, fraud interception and post-fraud feedback.
Anti-Fraud Algorithms: The company filters prospective borrowers through the use of anti-fraud algorithms based on machine learning: it utilizes supervised machine learning processes to learn from known fraud behavior patterns, training its algorithms to develop rules to identify similar patterns and deny suspicious applications; it utilizes unsupervised machine learning to run anomaly detection to detect individual and aggregated abnormal patterns for the purpose of identifying unknown fraud behaviors; and it conducts a social network analysis, connecting seemingly unrelated factors to often detected fraud schemes. For example, when a new user uses the same mobile device as that of users A and B to access its services, its social network analysis algorithm is able to automatically catch the high correlations that may exist between the new user and the existing users A and B.
Proprietary Credit Scoring and Risk Models
When a credit application is deemed to not represent a fraud risk, it is then subjected to the credit assessment module of its Argus Engine. This module will select and analyze variables associated with a given credit application. The variables that the Argus Engine analyzes are selected based on the perceived risk profiles of the applicants. The Argus Engine ultimately generates an A-Score to quantify an applicant's credit profile. Prospective borrowers with higher A-Scores typically receive recommendation for higher credit limits. The A-Score is then directed to the Cosmic Cube Pricing Model for pricing.
The company conducts credit assessment each time a new borrower requests a drawdown. A-Score is the result of the initial credit assessment performed on an applicant based on credit his/her credit profile, considering various factors, such as financial condition, education, past credit history and social behaviors. Different from A-Score, B-Score is applied to existing borrowers on its platform with more than three months of borrowing history, by monitoring borrower behaviors, such as account, drawdown, repayment, among others. The B-Score replaces the A-Score for the purpose of future credit assessment and re-evaluation. The B-Score is reevaluated each time the borrower applies for a drawdown and at the end of each month. Given that the company has high repeat borrower contribution, B-Score, reflecting the latest borrower behavior, plays a relatively more prominent role in its overall credit assessment process.
Based on the B-Score assigned to borrowers, the system adjusts recommendation of their credit line both proactively and in response to the requests made by them. For a given borrower, the request for credit line adjustment can be done no more than once every three months. A typical 15% to 25% increase will be given to the credit line of the borrower if the underlying adjustment is approved.
Real-time Risk Events Monitoring
Leveraging the expansive and complicated relational network of a borrower's financial connections, Argus Engine can extract the most important information from the massive dataset and determine the borrower's credit profile. When a borrower makes an online credit drawdown or application, the company needs to conduct real-time credit assessment, which necessitates the support of a powerful credit profiling engine. As of December 31, 2024, the real-time graph engine was in the fourth generation with more than 3.0 billion nodes and 171 billion edges. It provides more than 152 million times online calculations daily, mapping first-degree connections in an average of 5 milliseconds, and second-degree connections in an average of 40 milliseconds. Backed by powerful computation, its real-time screening net can accurately identify risks from group fraud, multiple platforms borrowing and default, among others.
Collection
The C-Score processes data from historical collection efforts to automatically identify the most efficient channel for collection, including text messages, mobile app push notices, AI-initiated collection calls, human collection calls, emails or legal letters. The company also outsources its collection to third-party collection service providers, particularly after 60 days of delinquency. To fulfill the compliance requirements, the company has adopted and enforced comprehensive collection policies and procedures, including close monitoring of its third-party service providers, to ensure that all its collection practices, including in-house and third-party practices, are in compliance with current laws and regulations. First of all, all collection operations, either conducted by its in-house collection team or through third-party agencies, must be processed on its proprietarily developed online operation platform and call-out platform so that the company is able to track and perform full-angle inspection on the collection practices. Secondly, all borrower data are subject to a desensitization procedure before they are used for collection. The company's system enables a close-loop monitoring over the process of the collection exercise, from case categorization and the desensitization of delinquent borrowers' information to the dispatch of delinquency information to the collection team or third-party collection agencies, as the case may be, and the collection call initiation. It ensures that only the necessary and minimum amount of desensitized data are being used for collection and that no data are able to be saved locally. Thirdly, all manual collection calls, either initiated by its in-house collection team or by third-party agents, are recorded and transmitted to its inspection system for an AI + manual dual inspection procedure, where its AI models will perform automatic, preliminary analysis on the content of the collection conversation against the rules that the company sets, identifying the expressions that are suspected to be deviating from its rules, and its inspection team will then further investigate the cases and provide improvement advice. Fourthly, the company maintains real-time inspection on all collection operations. The company's system constantly analyzes the real-time recording of the collection calls for potential defects or violations. Once a defect or violation is identified, a notice will be promptly sent to the on-site collection supervisor for intervention, so that the company is able to proactively de-escalate the situation, prevent violative collections and deliver better user experience. Last but not least, the company stipulates into each service agreements with its third-party agencies obligations of such agencies to abide by its policies, comply with laws and regulations, preserve confidentiality, refrain from using excessive or otherwise inappropriate measures.
The company has built an AI-powered collection and borrower service system based on automatic speech recognition, text-to-speech and natural language processing technologies. In 2024, the application of its AI-powered collection had handled 72% of its total collection volume. The company's collection system can conduct automatic outbound calls in batches and interact with borrowers. The company assess the appropriateness of AI-driven communication and will adjust the approach and tone of the system, based on the risk level and the type of collection. This assessment is conducted automatically, and the company leverage the capability for all early-stage notification, contact confirmation and basic collection negotiations, while focusing its collection team on complicated collection cases, or other challenging interactions as identified by its system, to increase its operational efficiency and reduce its collection costs. In 2024, the company maintained a 30-day collection rate of approximately 86.7%.
Data And Privacy Protection
The company is dedicated to protecting users' privacy, and it has implemented a data privacy and security system to ensure the security, confidentiality and integrity of data. The company adopts policies to make sure it obtains users' consent in collecting and using their data. The company has promulgated a user privacy policy on its platform, setting forth its data use practices and privacy protection protocols. When a user registers an account via its app, he or she must read through and agree to the privacy agreement before the registration can be completed. Besides, in certain phases of the loan application process that involve data collection or usage, such as activating facial recognition function to facilitate credit assessment and transaction security, its users will be prompted again to read through and agree to separate authorization agreements on its data collection and use practices before they can proceed. The company only uses the data for the stated purpose as authorized by the user of its app in connection with credit assessment and as otherwise required by applicable laws and regulations. All data which the company collected and generated from its operations in the PRC are stored in the PRC territory and the data which the company recognize as sensitive data are encrypted with the double encryption approach of data encryption and database encryption. The company stores user data in accordance with applicable laws and regulations, and the company has adopted and implemented internal controls system and protocols focused on data security and personal information protection. The company's core systems have all passed and been certified as the Level III Protection of the National Information System: The company requires all of its employees to comply with the protocols, respect the privacy of users, and protect their information. In addition, the company limits its employees' access to de-identified information and the output of such credit analysis only (except for key data security personnel whose access is subject to stringent internal approval) for purposes of mitigating the possibility of data leakage and avoiding unnecessary privacy invasion as much as possible.
With rigorous data privacy and security system, in June 2020, its fintech service application, Qifu Jintao, which was previously known as 360 Jietiao, received both the app security certification and the app information security certification from the National Computer Virus Emergency Response Center, which is the official agency for anti-virus internet security and designated testing body for the Special Crackdown on the Illegal Collection and Misuse of Personal Information by Apps initiative by the Ministry of Public Security. In particular, Qifu Jietiao received a level 3 rating for both app privacy and data security, the highest level granted by the center. Given the ongoing regulatory environment, the certifications granted to the company recognizes its core competency in privacy protection and security technology and further solidify its competitive advantage in terms of regulatory compliance.
Technology and Security
The company is a technology-driven company. The success of its business is dependent upon its technological capabilities, which deliver a superior user experience, protect information on its platform, increase operational efficiency and facilitate continued innovation. The company's innovation efforts are driven by strong research and development and risk management teams, which accounted for 35% of its total employees, as of December 31, 2024.
Principal components of the company’s technology infrastructure include:
Data Science: Data science contributes to many elements of its business and operations, extending across an entire loan lifecycle. The company's Argus Engine allows the company to aggregate and assess thousands of data points to build a comprehensive profile for each user which guides fraud detection, credit assessment and general borrower behavior, useful in anticipating borrowers' needs. The company's workflow system CloudBank is capable of processing millions of transactions every day and integrates with its financial institution partners' systems in loan disbursements, credit decisions, and payment clearances. The company has also developed its network relationship database with tens of billions of connecting points for fraud detection purpose. The algorithms powering the majority of its decision systems iterate in real-time through machine learning, allowing the company to promptly identify and correct operational issues.
Artificial Intelligence: The company has identified specific applications for AI across its platform, notably around precision marketing, rapid underwriting and post-facilitation services. The company consistently upgrades its capabilities through machine learning. For instance, its fraud detection and credit assessment capabilities are based on the self-learning of the Argus Engine, which consistently re-evaluates statistically significant variables and re-develops policies around borrower credit assessment. A key benefit of AI is the automation of many of the company's processes. The company can generally process a credit application from submission through drawdown approval without material human intervention, and its internal preliminary credit assessment mostly only takes less than a minute, in accordance with recent IT records, achieving massive operational efficiency. For example, the company's AI-powered voice system, which it utilizes for the collection of delinquent loans, primarily conducts quality inspections on the recordings of collection calls, analyzes the effectiveness and content of completed collections, and optimizes future communication strategies has significantly reduced the need for collections staff, particularly those involved in quality inspection, and has empowered the remaining team to operate more efficiently and effectively.
Security: The company is committed to maintaining a secure online platform. The company's platform benefits from 360 Group's expertise in the area of internet security. Key features of its security system are as follows:
The company's firewall monitors and controls incoming and outgoing traffic 24 hours per day, and the firewall is updated and trained periodically with mimic attacks from hackers to spot potential loopholes and protect its platform from malware, computer virus and hackings;
The company's servers are managed by 360 Group's private cloud service and as such are both physically and virtually isolated with intensive security protocols; and
All transmission of borrower information is encrypted.
The company has also adopted a series of policies on internal controls over information systems and network access management. The company maintains redundancy through a real-time multi-layer data backup system to prevent loss of data resulting from unforeseen circumstances. The company conducts periodic reviews of its technology platform, identifying and correcting problems that may undermine its system security.
Stability: The company operates on 360 Group's private cloud. The company's system infrastructure is hosted in data centers at three separate locations in Beijing and Shanghai. The company maintain redundancy through a real-time multi-layer data backup system to ensure the reliability of its network. The company's platform adopts a modular architecture that consists of multiple connected components, each of which can be separately upgraded and replaced without compromising the functioning of other components. This makes its platform both highly reliable and scalable.
Scalability: With a modular architecture, its platform can be easily expanded as data storage requirements and user visits increase. In addition, load balancing technology helps the company improves the distribution of workloads across multiple computing components, optimizing resource utilization and minimizing response time. Meanwhile, the company has built its system in a partner-friendly approach as it provides flexible options to its partners regarding the scope of the data to be provided as well as how the data is provided.
Marketing and Brand Awareness
The company primarily employs and implements variable online sales and marketing methods, supplemented with traditional promotional activities and general brand and awareness building. The company focuses on building brand awareness through online marketing campaigns, including cooperating with leading online platforms for directing user traffic to its business and boosting public relations as well as other offline advertising. The company invest in a series of marketing activities to further solidify its brand image and continue to grow its user base, including collaborating with leading social media, video and live streaming platforms to extend its brand to a broader potential user group.
Seasonality
The company experiences seasonality in its business, mainly correlating to the seasonal fluctuations in internet usage and traditional personal behavior patterns in China. For example, individual borrowers generally reduce their borrowings during national holidays in China, particularly during the Chinese New Year holiday season in the first quarter of each year. Furthermore, when e-commerce platforms hold special promotional campaigns, for example, on November 11 and December 12 each year, the company typically observes an increase in borrowing proceeds immediately following these campaigns. However, the seasonal trends that the company has experienced in the past may not apply to, or be indicative of, its future operating results.
Intellectual Properties
As of December 31, 2024, the company had 310 registered trademarks and 105 trademarks pending approval in China, 445 registered patents and 406 patents pending approval in China. As of December 31, 2024, the company had 144 registered software copyrights and eight copyrights of works in China. The company is also the registered holder of 71 domain names in China.
Regulation
The company provides Credit-Tech services for which a VATS License is required. Shanghai Qiyu, one of the VIEs, obtained its ICP License, a type of VATS License, in April 2021. The subsidiary of Shanghai Qiyu, Fuzhou Microcredit, obtained an ICP License in April 2023.
The company has taken various measures to comply with Circular 141, the Interim Measures for Administration of Internet Loans Issued by Commercial Banks and other laws and regulations that are applicable to its loan facilitation business operations.
History
The company was founded in 2016. It was incorporated in the Cayman Islands in 2018. The company was formerly known as 360 Finance, Inc. and changed its name to 360 DigiTech, Inc. in 2020. Further, the company changed its name to Qifu Technology, Inc. in 2023 and then to Qfin Holdings, Inc. in July 2025.
