The cost of a simple payment question
It’s 9am and you’re checking up on today’s settlement. It’s different from what you received yesterday. The team hasn’t figured out what happened. You open the dashboard, find the settlements tab, filter and export the data and manually analyse everything.
Half an hour later, you realise that one batch was pushed back by a day due to a bank holiday.
Finding the answers to your payment queries isn’t hard. It just takes up somebody’s morning. And it doesn’t stop at settlements.
An hour later, someone else is back on the dashboard, looking up a customer’s refund details. They weren’t able to complete payment so now your teammate locates the transaction, reads its lifecycle, checks the balance & applicability before issuing the refund. The same process is repeated over and over again for all refund claims of the day.
When the month ends, reconciliation needs to be done. Your team pulls the data, matches the records, tracks down discrepancies and reworks it into a format your finance team uses. Then there’s always the sales unit that logs in multiple times a day. They create payment forms and links to share with customers as soon as a requirement comes in.
Our dashboard was built to store all the information collected about your business’s payments. Answers to your questions exist in one place but they hide behind tabs and filters and require you to do the reading and analysis on your own.
Even reports and links that are used everyday have to be created and all seem simple at surface level. But what happens when questions come in faster than your team can navigate? As your business scales, the questions snowball. Before you know it, you need answers to a hundred things simultaneously. The dashboard stops being the tool and becomes a bottleneck.
Most tools address this problem by adding another screen, making it easier to find the answer. Spark gives you the answer.
What’s Spark?
Spark is the agentic command surface for your business. Built into the Cashfree merchant dashboard, it lives in the search bar at the top of every page. It turns the dashboard from a space that holds payment information to one that finds and shares answers.
Ask Spark on chat or speak in your preferred language. It reads your Cashfree account data and answers instantly with real figures and the records they came from. If it’s in your merchant account, Spark can reach it irrespective of whether it’s payments, refunds, settlements, disputes, reports, links, offers, webhooks, preferences or anything else. You don’t have to use a filter or install anything new.
Getting answers is the easy part. What happens next? Unlike generic AI assistants with limited capabilities, Spark can act on behalf of your team. With one command, it can create payment links, put together reconciliation reports, process refunds, and initiate instant settlements. Tasks that took up to half a day are completed and shared with your team in minutes.
How does Spark work?
There are two things that Spark will never do: answer without real facts & figures and move money without your approval. Let’s take a look at how Spark thinks and works. Before answering, it tries to understand the context. Whether you ask Spark questions or give it a task to carry out, it reads into your requirement and identifies what to look for.
Each answer is assembled from real calls to your live Cashfree system. Spark uses its tools to search through your account and shares data backed from your dashboard. It doesn’t answer from memory so for every question you ask, a real call is made. When you ask for the settlement amount at 4pm today, it gives you the updated amount at that moment. And if Spark can’t find what you’re looking for, it tells you that directly instead of taking a wild guess.
Any action that requires Spark to move money happens only after you approve. Spark draws a hard line between answering and acting. Whether it’s refunds, a payment link, offers, settlements or changing the current setup, it asks for your permission every time. It states the exact action and values for you to review and only proceeds after you sign off.
Spark works within the permissions you set on your dashboard. Before a task is carried out, it checks the authorisation of the teammate requesting the action. If they can’t complete the action manually on the dashboard due to the permissions set, they will not be able to get it done through Spark either. Permissions will have to be updated.
It only happens with your sign off irrespective of how big or small the task is. This check happens every time a request is made. It’s based on who logged in and what their role allows. All the details are recorded in your account’s audit trail.
Each ask gets an agent of its own
One reason Spark finds exactly what you want it to find is because it has a team of specialists to tackle payment queries. Spark routes each question to its respective agent. Settlements, refunds, offers & checkout, disputes webhooks and error logs and reports all have an agent with a unique set of tools. This way, your settlement agent knows how your settlements work and your refund agent knows everything about refund eligibility.
These agents read your questions, calls the tools it needs, finds your answer and either share that with you or ask you a follow up question. Along with the result, they also suggest the next steps so a look up can be followed by its succeeding action. The mechanics of Spark are not something that is evident on the surface. But this is precisely what enables Spark to give accurate answers for settlements and webhooks and all your other queries.
What can you ask Spark?
Describe what you want in plain words in your preferred language. Spark understands this better than complex commands.
Here are a few examples to begin:
- Summarise my sales today.
- Why is today’s settlement lower than yesterday’s?
- Withdraw what I can instantly settle. Show me the fee first.
- Which alerts to my system failed yesterday? Send them again.
- Create a 10% discount offer for all customers.
- Build a reconciliation report for the last month.
- मेरा अगला सेटलमेंट कब आएगा
Spark reads your request and figures out the look up needed to answer or carry out the task.
| Area | Your Request | What Spark does |
|---|---|---|
| Settlements | Check my Cashflow | Explains a short settlement, shows what is eligible for early withdrawal with the fee first, initiates instant settlement |
| Refunds | Clear my refund queue the same day | Checks eligibility and balance, then issues the refund |
| Reports | Carry out the reconciliation for the last month | Builds reconciliation and summary reports on command, then returns the download |
| Payments & orders | Why did Rohan’s payment fail? | Finds the transaction & explains failure reason |
| Links & offers | Ensure that live offers don’t clash and my customers receive unique links | Checks active offers, pauses conflicting ones, creates unique payment links |
| Webhooks & errors | Which webhooks failed yesterday? | Finds what failed, replays it & explains the error |
| Preferences & whitelisting | Add my new domain to the whitelist and increase my limits. | Changes settings, submits domains, adjusts limits |
| Product docs | How do instant settlements work? | Explains how this Cashfree feature works. |
Your team gets back to growing the business
Payment operations no longer depend on your team’s headcount. The hours spent everyday checking up on settlements, screening refunds and building reports can all go back to growing your business. Customer queries get answered within minutes as payment failure reasons are now shared instantly.
Reconciliation reports are automatically formatted to what the finance team follows. Your sales team now sends payment links to customers while their intent to buy is high. The volume of questions surely doubles as your business grows but answering them will remain effortless.
Just ask Spark and it’s done.
