Recko is a fintech startup that enables AI-powered reconciliation of digital transactions. Running on a SaaS model, the platform helps in monitoring and reconciling all digital transactions throughout the transaction lifecycle. This process eliminates the tedious and fallible process of manually tracking payments from the customer to the seller through various payment nodes. An independent third party transaction reconciliation layer such as Recko ensures that each transaction is accounted for and all settlements can be done in a timely manner.
Founded in 2017, the startup is at a run rate of reconciling a quarter billion transactions annually. It has reconciled transactions worth $2 billion in the first 12 months of operations. Their clientele includes companies such as Grofers and Meesho. Recko's solution is suited to any company with high volumes of transactions such as e-commerce sites, insurance providers and banks.
Recko has recently raised $1 million in seed funding from Prime Venture Partners. After raising their seed fund, the company will now focus on expanding their team, scaling operations and launching into unexplored sectors.
CEO and Co-Founder Saurya spoke to TNM on the inception of the company, challenges in building the product, expansion plans among other things.
How did you come across the problem?
As we started scaling up, a lot of merchants came on board. So, we had about 10000-20000 merchants on the platform. Similarly, on the consumer side also, the number of orders were increasing very rapidly. And whenever these orders are there, you have payments. Now, there were sellers who were selling on our platform and once we get the money from the customer, then we have to settle it to our merchants because these are the people who are selling the product. As volumes started shooting up, it became difficult for the finance team to keep track of the money movement.
People were dealing with complexity and good amount of transactional volumes but they did not have the tools to take care of this data. The finance guys only have an Excel worksheet and a laptop. Excel has a limitation of a million records. Moreover, these transactions are running transactions.
The finance person has to ensure that the money which got processed by the payment gateways is hitting the bank account and against the same set of transactions, there are refunds which are getting processed; there are exchanges which are in the supply chain. At any point of time, the internal order management system and the bank account have to be in sync. And this has been a challenge for almost everyone who has been there in the ecosystem majorly because of volume and complexity. That is where we come into the picture and where we help the finance departments of internet companies in handling this transactional volume.
We started in May 2017 and for the initial few months, we just worked on the product. Then we onboarded our first client and took on more clients. It took us some time to stabilise the product because the volumes that we were catering to, were unusually high. About Rs 20-30 million used to hit our system at one go. But during that time, the product got built for scale. Now we donâ€™t have any problem even if somebody pushes 100 million items at one go. We are pretty settled now to handle any kind of complexity.
Challenges in building this product
The complexity was very high. So if you move from say company to company, at times companies have a different structure of handling this data. A food tech company might be different, a transportation company might be different. So building a generic product for them takes a lot of time, which is where a lot of our investment and time went in.
Since we were majorly dealing with really high data volumes, the product had to be stable. The product is constantly evolving as we add features to it. The idea is that as you move forward, the number of bugs, the number of inconsistencies keep on reducing.
So the challenge which we had since it was a financial software was that we had to be accurate and we had to be accurate at scale. We spent a lot of time in making sure that the numbers are exact.
On the engineering side, it was a huge challenge for everybody inside as to how do we process these transactions because it used to take us more than 3-4 days to process more than 50-60 million transactions. Now, we are able to do it in 30 minutes. Similarly, for reconciliation, we are almost running at 100 million transactions in one hour. So the systems are becoming much faster. The idea is how do we do this at a much cheaper cost and faster. So this is where a lot of investment is going in.
From a data security point of view, we have built controls inside the product so that it functions in the way it has been designed and there is no deviation from that.
We are also completing our ISO and PCI certification. As we move forward, we will also be completing SOC (Service Organisation Control) 1 and SOC 2 which are gold standard when it comes to internal data.
What are the expansion plans?
Right now, we are focussed on the India market and the idea is to essentially cover certain verticals as fast as we can, probably in the coming few quarters. After that, we would be most likely going outside. We already have a lot of inbounds which are coming from Middle East Asian countries, Southeast Asia. But right now we want to keep our focus on India and complete the milestones which we have kept for ourselves.
Within India, we are focussed on the internet companies because we have built a lot of these systems in our previous roles. So we understand the internal systems very well like order management system, supply chain, payment, accounting.
Some of your major clients
Meesho and Grofers have been running on us. There are a lot of marketplaces which have been running on the system. As we are moving forward, we have also started working closely with a few banks, a few public insurance companies and also couple of companies from Africa, Middle East and Southeast Asia.
How do you see this market evolving?
The nature of problem is going to become exponentially complex because so many layers are getting created on the payment side. People keep on integrating their SDKs (software development kit) into somebody elseâ€™s. So the moment that happens, there is one more layer of reconciliation which gets created. As people are creating more layers, the number of reconciliations will just shoot up.
We are building a lot of capabilities on the data science and ML side, which can help people in identifying these trends and ensuring that you donâ€™t settle someone twice, donâ€™t receive double payments and donâ€™t end up paying more refunds.