[Product Roadmap] With Grofers, Dunzo, Udaan as clients, how fintech startup Recko uses tech to simplify reconciliation
India has been seeing a surge in digital transactions in the past few years. Most organisations in the country are constantly grappling with increasing transaction volumes and complex payment flows and systems.
Entrepreneurs Saurya Prakash Singh and Prashant Borde realised one major problem faced by finance teams in organisations, i.e., reconciliations. It got them to start Recko in 2017.
Today, the fintech startup is helping customers across industries like banks, ecommerce, food tech, fintech, neo-banks, gaming, insurance providers, etc., which have a high volume of transactions, by automating their reconciliation process.
“We work with some of India’s top companies like Grofers, Meesho, Dunzo, Pharmeasy, Curefit, Udaan, MPL, Khatabook, and others,” says Prashant.
Solving a complex problem
Prashant says most organisations receive data from multiple sources such as settlement files, order management systems, banks, etc. These files are in different formats and they need to be cross-checked to ensure the information available across is in tandem with one another.
The finance teams have to transform these data and then reconcile them. More data processing is required when working with different finance tools used for accounting, tax, payouts, etc.
“One can only imagine how difficult things would turn out to be when the volume increases,” says Prashant.
He says Recko is solving this by using Big Data frameworks, which are tuned for precision, consistency, and scalability. The startup identifies the data that is fed into the system with the help of pre-defined rules and reconciles at a transactional level.
The platform provides a comprehensive and robust technology stack to manage financial data and enable financial workflows such as reconciliation, commission calculation, payout creation, and reporting for businesses to track, manage, and account money end-to-end.
Understanding the deeper problem
Explaining the challenges in the financial systems, Prashant says, merchants face issues when it comes to reconciling these payments due to outdated tools and technology.
At present, the whole process of ensuring if the amount in the settlement files and bank account are in agreement with each other is done manually by most companies.
“We started thinking what if there is a platform that can enable businesses to reconcile easily. Saurya and I started talking to finance teams in companies across different verticals and geographies (India, the US, SEA, etc). They all faced a similar problem and were using Excel sheets to reconcile payments,” says Prashant.
The duo also understood that tracking commissions were another unresolved use case. “No tool offered this flexibility without getting too deep into tech,” says Prashant.
They aimed to automate and fast-track this process to help businesses track accurate revenue, close SLAs for receiving the money to optimise working capital, and reduce support cost by having a clear picture of money paid to partners.
Image and Data - Recko
The first working prototype of the product (codename Perlis) was built by Saurya and Prashant with the help of a few developers. “It took us a few months. Saurya used to take care of the product, design, and customer communication aspects and I focused on the tech,” says Prashant.
“We focused a lot on the accuracy, which helped us gain the trust of the finance teams and gave us the confidence to march further,” he adds.
However, one of the challenges they faced was supporting scale. Prashant says, they had to rewrite a large part of the first version to make scale and security a foundation in the architecture.
“We then released the next version of our product (codename Wilkes), which supported scale, accuracy, and security,” says Prashant. Wilkes eliminated most of the operational work and they started onboarding more customers onto the product.
With Wilkes, customers were able to automate and increase the speed of their monthly activities, which helped them save about 70 to 80 percent of their time, he says.
“We continued working on the enhancements and hired a full-time UX designer to improve the intuitiveness of the product. Users wanted more out of the product and we started enabling them to automate tricky use cases. They needed more analytics and alerting capabilities. We also figured out that manual push of data affects data quality and we need direct integration with data sources,” says Prashant.
Some of the early adopters of Recko’s products were in the ecommerce segment. Prashant says they onboarded giants like Myntra, Meesho, Dunzo, PharmEasy, and Grofers, who helped them expand into the on-demand food tech industry.
Building a global product
“As we onboarded new customers, we realised that businesses looked at data very differently across industries. We did not want to leave any stone unturned, but we had a mission — to give the best of it. We added analytics, custom reports, commission calculation, and other integrations including storage services, payment gateways, and banks,” says Prashant.
The team soon started working with customers from different industries and locations. Recko added geographies like SEA and the EU. They also released versioning to support audit logs and time travel features that required an overhaul to support future scale.
The team then realised it needed to adopt new technologies like Spark, Data Lake with rollback support, and Kubernetes, to help ease with volumes. For this, Recko had to rewrite a large part of the codebase to address tech debt and make it more modular. Prashant explains,
“Considering the global outlook, we needed a design overhaul as well. So, we started working on the next version (codename Minsky). We discussed state machines, flowcharts, and architecture diagrams, and once the initial draft was ready, we ran it through a few industry experts to avoid common pitfalls."
"Until this version, we relied on fast execution to incorporate and address tech debts through revamps. After this version, we were able to make the product more modular and continue product development without much technical tech debt.”
Building more distributed systems
Prashant says the team then focused on the reliability and extensibility of the system and there was a shift in the tech thinking process itself.
Today, the fintech space of reconciliation is fast growing with startups like PagarBook and others focussed on the growing SME segment. Recko, on the other hand, is sector-agnostic.
“We were thinking of a more distributed process, trying to familiarise no-SQL design patterns using seamless, columnar databases,” says Prashant.
Recko is now planning to add more modules like payouts calculation and scalable ledger. The payouts calculation module will help startups and mid-market companies have payouts set up without investing any time in its development. And the ledger module acts as a single source of truth for financial data.
“We are planning to open APIs as well so that they can be integrated deeper into companies’ tech stack to solve a multitude of problems. Our long-term goal is to provide enough insights that enable businesses to make financial decisions in real-time,” says Prashant.