Deposits approve themselves. Your team only sees the exceptions.
Loop Trade reads your payment gateways, matches each deposit by UTR and amount, and approves it in seconds.
Pending request
Vikram T.
UTR 4021 8873 1592
₹24,500
Gateway feed
Google Pay
4021 8873 1592
₹24,500Bank (IMPS)
7710 4438 2201
₹8,000Bank (UPI)
3390 2216 8874
₹1,25,000
2.5s
Match loop
30s
Capture
24×7
Uptime
A demo run of the matching engine. Amounts and references are sample data.
Reconciles every rail your desk takes money on
Manual reconciliation fails at the worst possible moment.
Deposits land in one system. The approval queue lives in another. Keeping the two in agreement takes a person, two screens, and all day.
Someone has to watch the gateways
Payments arrive in Google Pay and your bank portals. The queue waiting to be approved is somewhere else entirely.
Peak hours are late at night
Requests cluster when the market is busy, which is exactly when staff are tired and error rates climb.
Matching by eye is fragile
Twelve-digit UTRs, near-identical amounts, the same sender paying twice. A transposed digit is invisible on a tired screen.
Reused references slip through
The same UTR submitted again against a different amount is a classic abuse pattern, and it is very hard to catch by hand.
Nobody sees the whole picture
How much came in today, how much is still pending, which gateway is underperforming. The answers arrive hours late, assembled by hand.
Rejections stall
An unmatched request sits in limbo until someone remembers to decline it, leaving the client and the desk both waiting.
Five stages, always running.
There is no manual step in the happy path. Each stage runs independently, so a slow gateway page never delays an approval.
Read every gateway
Capture
Loop Trade continuously reads the transaction lists from each configured gateway: Google Pay and your bank portals.
- Extracts UTR, amount, sender name and sender detail from every row
- Runs many gateways in parallel across a small pool of browser tabs
- Keeps one persistent, logged-in session per site
- Separates pools by rail, so a UPI payment is never reconciled against a bank transfer
The whole operation on one screen.
Today's numbers by default, streaming in as events happen. No refresh, no spreadsheet, no waiting for a shift report.
Requests today
1,284
+6.2%
Approved
1,171
91.2% auto
Pending
42
held for review
Rejected
71
no payment found
- Approved
Approved automatically
Google Pay · UTR 4021 8873 1592
₹24,500
Rohit S.
- Approved
Approved automatically
Bank (IMPS) · UTR 7710 4438 2201
₹8,000
Ananya M.
- Held
Held for review
Same UTR seen with a different amount
₹1,25,000
Imran K.
- Approved
Approved automatically
Google Pay · UTR 9128 0043 7712
₹4,999
Priya N.
- Rejected
Auto-rejected on timeout
No matching payment within 5 minutes
₹60,000
Sana R.
Illustrative preview of the live dashboard.
Everything the desk was doing by hand.
Capture, matching, approval, sync and reporting, in one engine that runs whether or not anyone is watching.
Two keys, one decision
A payment is matched only when the UTR and the amount agree. Amount alone is never sufficient evidence.
UTR
4021 8873 1592
Amount
₹24,500
Both agree, so the match is confirmed
~2.5s matching loop
The matching engine runs on its own tight loop, deliberately separated from the browsers.
2.5s
Parallel capture
Many gateways read at once across a small pool of tabs, each with its own persistent session.
Approve and reject
Verified matches approve instantly. Unmatched requests time out on their own, so nothing stalls.
Verify before persist
The platform is checked before the database is written, so a record never says approved when it is not.
Deposits and withdrawals
The same engine drives withdrawal processing, with the same rules and the same audit trail.
Two-way platform sync
Decisions made here are mirrored there, decisions made there are read back here. The trading platform stays the source of truth.
Human in the loop
Reused, mismatched and suspicious entries arrive pre-loaded with history for a reviewer.
Two keys, not one.
Most mismatches come from trusting a single signal. Loop Trade refuses to.
- Pools stay separate. UPI is never cross-matched against IMPS or NEFT, even at an identical amount.
- Nothing is guessed. Ambiguous or reused entries are escalated with their history.
- Timeout clocks are anchored to stored timestamps, so a restart loses no in-flight state.
- Every loop iteration is guarded. One bad entry is skipped and the pipeline continues.
- Rohit S.4021 8873 1592₹24,500
- Neha D.5580 1129 4470₹24,500
The exceptions that deserve a human.
Automation should not mean blindness. Five cases are deliberately escalated, each arriving with the context already assembled.
Why it is held
The same UTR appears again against a different amount, which is a classic reuse pattern.
What the engine does
Held for review, then auto-rejected if it times out.
History attached to the row
- 12 Aug₹18,000Approved against a Google Pay creditApproved
- 23 Aug₹42,000Same reference, a much larger sumHeld
The numbers the desk cares about.
Reported the way the system is designed and observed to behave. Confirm current figures against your own production data.
Zero-touch resolution
90%+
Deposit requests resolved with no human involvement at all.
Matching loop
2.5s
Continuous comparison of payments against the pending queue.
Gateway capture
30s
Per gateway, read in parallel rather than one after another.
Unattended
24×7
Self-healing and supervised, with no per-transaction fee.
Built to run for months without supervision.
The reliability work is not a feature list, it is what makes a 24 by 7 operation possible without a person on standby.
Supervised processes
Every component restarts automatically on failure, with backoff so a bad state cannot become a restart loop.
Self-healing watchdog
Each minute it checks the display stack, the browser, the web app and every worker, and repairs whatever is down.
Memory watchdogs
Node and browser memory are watched and cycled before they become a problem, not after.
Circuit breakers
Repeated gateway failures trigger a cooldown instead of a spin loop that burns the machine.
Persistent sessions
Logins survive restarts. The browser is kept alive and reconnected to, never relaunched from cold.
Scheduled self-restart
A daily restart at a quiet hour, with old logs and notifications pruned automatically.
The full checklist, including data handling and audit, is on the security and reliability page.
What the manual queue costs you.
Move the two sliders. The estimate assumes a human check per request, which is what the automated path removes.
Resolved with no human touch
720
requests per day, at a 90 percent zero-touch rate
Held for a human instead
80
the genuine exceptions, reviewed with history attached
Manual checking removed
24 hper day
about 720 hours a month
Still manual
2.7 h
per day
An estimate based on your own inputs, not a measured result. The demo is where we check it against real numbers.
Questions the desk usually asks.
If yours is not here, the demo is the fastest way to get a straight answer.
- What does Loop Trade actually connect to?
- Google Pay and your bank portals over UPI, IMPS and NEFT, plus your trading platform for two-way sync. Any gateway you can open in a logged-in browser session can be configured as a source.
- Does it approve payments without a human?
- Yes, for roughly nine in ten requests, and only when a payment matching both the UTR and the amount is sitting in a gateway. Anything ambiguous is held for a person instead of guessed at.
- What happens to a request that never matches?
- It stays pending until its configured timeout expires, then it is rejected automatically with the reason you set. The timer is anchored to a stored timestamp, so a restart does not reset it.
- How does it handle a reused or duplicate UTR?
- The entry is flagged and the full history is attached: when the reference was used before, for how much, by whom, and how it was resolved. A reviewer decides in seconds rather than investigating in minutes.
- Where does my data live?
- On infrastructure you control. The application, the database and the browser sessions all run on your server, and a deploy never disturbs a live session because browser profiles are kept outside the application directory.
- What happens when a gateway page changes?
- Capture is isolated from matching, so a broken page stops new payments from arriving but never corrupts a match already in progress. Repeated failures trip a circuit breaker and raise an alert instead of retrying forever.
- Will it fight with decisions my team makes on the platform?
- No. The trading platform is treated as the source of truth in both directions. A decision made there is read back into the dashboard, and a decision made in the dashboard is mirrored onto the platform.
- How is it priced?
- Per deployment, scoped to your gateway count and daily volume. There are no per-transaction fees. Tell us your setup in the demo and we will quote it directly.
Watch it clear your queue.
Bring a day of real deposit activity to the demo. We will run it through the engine and show you exactly what it resolves on its own.