In 2016,was started by Shruti Aggarwal to make credit loans easy, transparent, safe and secure for borrowers across India. The startup serves a customer segment largely ignored by banks by offering them a product that would suit a bank’s premium customers. These can be students, the unbanked, and new credit consumers. Parikshit Chitalkar and Tushar Aggarwal are co-founders of StashFin.
“We are targeting a wide range of customers for whom we need to create underwriting models to support them, and have operations that can serve those customers very effectively. It requires a lot of rapid change because we don’t really have another competitive product that we can learn from, ”says Shruti. Your story.
On StashFin, personal loans are issued either via a bank account or via an innovative prepaid Visa card, with interest charged only on the amount withdrawn, making it suitable for borrowers who want financial flexibility at a lower cost.
“The nature of our business is very similar to what a bank would do. So imagine that every application or solution that a bank manages with thousands of people in technology is created and managed by us with less than 40 people, ”says Shruti.
The technology covers the entire customer lifecycle – from origination, onboarding, decision making, loan management, repayments, dashboards, reconciliation, loan operations and risk analysis.
Complete technical control
“We rely heavily on machine learning and use it to optimize various aspects of the business, from risk assessment and decision making to optimizing operational efficiency. The stack has helped us go from processing around 100 loans per month initially to over 150 loans per hour, ”she adds.
AWS recently featured StashFin as one of the top fintech stacks in India after extensive platform review. During these four years, StashFin has seen nearly 800% growth in user numbers.
Speaking about the importance of a full technical check, Shruti says, “For example, we need to have server-side content delivery in our application so that if we decide to open a low-income customer segment to Odisha, we can quickly deploy a targeted page with different images and different linguistic localization, while still having a well-to-do Gurgugram client logging into the same application.
Thus, Android / iOS versions should be carefully planned with updates to be minimized by performing configuration downloads at runtime rather than releasing new resources.
One of StashFin’s flagship products was conceptualized and launched on the app in less than five days. This product now makes one loan per minute and attracts over 40,000 new customers each month.
“Another example is that during the lockdown, our infrastructure team was able to set up the Work-at-Home Infrastructure (WFH) in less than eight hours, so every application available in the office worked on home networks, including VOIP calls. for our call center executives. Our professional call center partners took weeks to get online, preparing our in-house team overnight was very helpful in the collections as we were able to call clients as we normally did from the office ”Says Shruti.
Build the MVP
For the Minimum Viable Product (MVP), the team wanted to focus on their time-to-market strategy, release the product quickly, learn from feedback, and constantly improve.
“We sometimes did more than 10 production versions in a day. The main goal of our engineering team is to develop a thorough understanding of the business. All of our developers understand the business in and out. They are collocated with the sales teams on the same floor, which automatically drives the product from the bottom up, ”explains Shruti.
Emphasis is placed on training the engineering team to be “Full-cycle developers” – from design to finding suppliers, including the call team and listening to customer phone calls.
Shruti explains that if engineers hear customer feedback firsthand, the product turns out to be better. She says the team removed the middle layers so reaction times were faster and the end result was a better product for the customer.
Technology is built in-house. Shruti says this made the hiring process difficult as they were looking for engineers who had good development skills but were also passionate learners at the same time.
“We are very attached to surveillance; we want to know what fails, where and why. This is the main source of our learning, then we train very strongly internally, do a lot of POC on new technologies and when something sticks, we acquire this skill and put it into production, ”she adds.
Keep it simple
StashFin is built on a cloud stackSo, she says, scaling the infrastructure was less difficult. The DevOps team and infrastructure have built tons of automation to keep engineers focused on the product. Everything from CI-CD pipelines and testing to quality reviews and code mergers is fully automated.
“We’re in a mode of continuous improvement – whether it’s product features, infrastructure improvements, or better engineering, we’re always doing new things. We’re very careful about putting experimental stuff into production, but we’re experimenting a lot, ”says Shruti.
Consumer apps were launched within three months of their creation. Approximately 7,000 to 10,000 users connect to it at any time of the day, and applications have accumulated more than 20 million downloads. On Google Play, StashFin has a 4.9 star rating.
“Each engineer has set alerts on their code in their modules that let them know how things are working. This way we get to know how people use the product. We found a lot of cases where users were doing something completely different than what we envisioned and we fixed it and then re-measured to see the gains, ”she explains.
Following user feedback
Another example is the addition of a vernacular function. The team found that users in Level III cities took longer to complete payments than in Level I cities. So they added a language-only option on this page and saw a huge increase.
“We put the responsibility on the team. We have two sessions per week for learning and development where people present what they have learned and tried during the week. We use these sessions to jointly review and improve the code, make the data model more efficient, optimize queries, and more. Adds Shruti.
She adds that this way they have more 30 people think about the same problem in different ways. They also invite business users to these sessions and often give them good feedback.
“We have to innovate to remain the market leaders and at the same time we are looking to make money, so running a stable system is absolutely essential. This presents a unique challenge for us. It is more difficult for us to effect large-scale changes very quickly, which can jeopardize stability. The way we get around this challenge is to automate, monitor, and alert strongly. In this way, we increase the team’s confidence in the system so that they can focus on innovation while the assurance of stability is handled by the machine, ”says Shruti.
The old belief that code can fix everything doesn’t apply to StashFin. The team now thinks less code is better. Shruti adds that they use managed services and have built integrations around them to get the job done. This greatly increases stability and the team is assured of performance on a large scale.
Speaking of the future product roadmap, Shruti says:
“We are working on several feature additions to our product. We are working on many integrations to support new partnerships. We have a solid roadmap of what needs to be developed by our machine learning team, where all of our models will become fully self-reinforcing. We are also constantly improving our fraud detection, where we are starting to experiment with NLP to analyze customer interactions with our executives during voice / chat calls, and find models to improve customer service and eliminate bad actors.