Why RakamHQ

The money nobody else can see.

The payment leg - gateway to wallet to settlement to fee to bank - is a black box to every dashboard and every accounting tool. That black box is our entire company.

Why RakamHQ

“But doesn't ___ already do that?”

No. We are not a gateway, not accounting software, not a store system. We are the one layer that proves the money - the part everyone else treats as a black box.

Excel + your accounts person
That is our real competitor. We sell their mornings back - fewer errors, a full audit trail, a fraction of the hours.
eSewa / Khalti dashboards
Each shows its own island. You live on four islands plus a bank. We are the map.
Tigg / accounting software
Tigg records the books; we prove the money.Our clean, matched entries become their input - we feed your accounting, we don't replace it.
Indian recon tools
They prove the category - and will never read a Bikram Sambat eSewa report, model wallet pooling, or speak a Nepali accountant's VAT reality. Localization is the moat.
L0
Excel by hand
L1
Books vs bank
L2
Provider reports
L3 · RakamHQ
Order ↔ gateway ↔ settlement ↔ bank, with fees & refunds
L4 · RakamHQ
Ask your money anything
The problem

Yesterday's sales is really four lists that never agree.

Your orders say one thing. The gateway panels say another. Money pools in wallets. The bank shows cryptic lumps, minus fees nobody explained. Reconciling them by hand in Excel is where every evening goes - and where money quietly hides.

List A

Your orders

Created100
Marked paid97
Proven paid?
List B

Gateway panels

eSewa success54
Khalti success34
Abandoned / dup6
List C

Wallet balances

In eSewa wallet84,300
Swept to bankno
Age6 days
List D

Bank statement

"ESEWA STL"?
Minus fees?
Which orders?

A ≠ B is normal. B ≠ C ≠ D is where money hides. RakamHQ is the map across all of them.

The levels ladder

Everyone stops before the money.

Reconciliation has levels. Most tools in Nepal live at L0 or L1. The value - and the difficulty - is in L3 and L4, where the payment leg finally becomes visible.

L0Excel by hand

Three portals and a spreadsheet, every evening. Slow, error-prone, no audit trail.

L1Books vs bank

Accounting software matches bank lines to entries someone already typed. One hop.

L2Provider reports

Ingesting eSewa/Khalti reports into a sheet. Better - but nobody does it well in Nepal.

L3Order ↔ gateway ↔ settlement ↔ bankRakamHQ

Cross-source matching, settlement decomposition, fees, pooling, netting, refunds. RakamHQ.

L4Ask your money anythingRakamHQ

A read-only AI over the proven ledger. Conversational, agentic money intelligence. RakamHQ.

The one-question test of the gap: “Which orders are inside yesterday's eSewa settlement, and what did the fee come to?”

Why this is hard to copy

The shallow version is easy. The deep version is a moat.

It's fair to ask why a wallet, a gateway, or an accounting suite couldn't just build this. They could try - here's why it's slower and harder than it looks.

Parsing is the doorway, not the building

Anyone can teach a statement importer to swallow a wallet report - that's a weekend of work. The moat is everything after extraction: matching across four independent sources, decomposing settlement lumps, auditing fees, and tracking pooling, netting and refund lifecycles. Extraction is roughly 10% of the pipeline. We permanently over-invest in the other 90%.

We manufacture the proven receipts

Accounting tools reconcile bank lines against entries someone already recorded - they presuppose the receipts exist. We produce them, exploding a bank lump back into the gateway transactions and orders it contains: data their schema has no tables for. That's why our output is literally their input, and why we feed accounting rather than compete with it.

The unglamorous middle is a second company

Per-bank parsers, netting states, exception operations, managed refunds, provider-terms compliance, courtesy rate budgets - the grinding, unglamorous work a dashboard or a books-first team's roadmap gravity avoids. Building it means running a second company while defending an existing product line. Buying it takes a quarter; building it takes years.

The AI sits on top, never inside the math

The ledger computes; the model asks read-only tools, then explains. Numbers never pass through the model, so it cannot hallucinate them. You get an exact, evidence-backed answer to 'where is my money?' with the rows attached - a conversational money layer, not a chatbot bolted onto a dashboard.

Localization is the moat outsiders won't cross

Bikram Sambat dates, wallet pooling, Khalti netting, VAT context, and a morning message in Roman Nepali. These aren't features an international recon tool bolts on - they are the whole product, built by people who text the way their customers do. An Indian tool will prove the category and never read a BS-dated eSewa report.

We're already inside the workflow

By integration design, our clean matched entries flow into the accounting a business already uses. At any point of evaluation, their books are already truer because of us - through their own tools. Removing us makes a customer's numbers worse. That's the strongest position a layer can hold.

Get your free reconciliation.

Send us last month's eSewa report, Khalti export, bank statement and orders sheet. Within 48 hours we send back a one-page close - every payment matched, every fee computed, every settlement decomposed.

Files only · private upload link · nothing to install