Reconciliation
Bank to ledger, GSTR-2B to purchase register, marketplace settlements to orders, payment gateway to invoices. Fuzzy matches are scored and explained.
Build
The back-office work, done by Monday morning.
Agents that reconcile, check, match, process and report across Tally, SAP, Zoho, your bank portals and your spreadsheets. They do the routine steps, flag the exceptions, and wait for a person to approve anything that matters.
- Typical length
- 3 to 8 weeks
- Indicative price
- ₹4 lakh to ₹20 lakh, excluding GST
- Engagement path
- Discover→Prove→Build→Run
Sound familiar?
Typical situations, not quotes from clients.
- Month-end close takes eight days. Four of them are matching bank entries by hand.
- Our team downloads GSTR-2B, matches it against purchases in Excel, and chases vendors by email. Every month.
- We bought an RPA tool. It broke every time a portal changed a button.
Rule-based automation works when the inputs are tidy. Indian back offices are not tidy: narrations that say “NEFT/ABC TRADERS/INV 4471”, vendor names spelt five ways, emails with orders typed in the body. Agents can read the mess, apply your rules, and leave a clear trail.
What we build
GST and compliance checks
Missing or mismatched GSTINs, input tax credit at risk, invoices not yet filed by vendors, e-invoice and e-way bill details that do not match.
Order and request processing
Orders arriving by email, WhatsApp or PDF are read, checked against stock and price lists, and entered into your ERP for approval.
Reports that write themselves
Daily MIS, ageing reports, exception summaries and vendor follow-up emails, drafted from live data and sent after a person checks them.
Approvals where they belong
You decide which steps run alone and which need sign-off. Postings above a threshold always wait for a person.
A trail for auditors
Every step logged: what the agent read, what it decided, why, and who approved.
How it fits together
Where it fits
- Tally Prime
- SAP S/4HANA and SAP Business One
- Zoho Books and Zoho Inventory
- Oracle NetSuite
- Microsoft Dynamics
- GST portal exports
- Bank statement files and APIs
- Amazon, Flipkart and other marketplace reports
- Shared mailboxes
- Excel and Google Sheets
Where a system has no API, we use exports, scheduled files or supervised browser steps, and we tell you the trade-offs of each.
How the engagement runs
Typical length: 3 to 8 weeks from pilot start to production.
Discover
1 weekWe sit with the team, time each step of the process, and collect three months of real inputs and outputs.
Prove
2 to 4 weeksA fixed-price pilot on one process, run in parallel with your team on live data. Scored against your team’s own results for the same period.
Build
1 to 4 weeksProduction integration, approval flows, scheduling, security review and team training.
Run
monthlyWeekly exception review, new rules added, accuracy reported against the agreed metric.
How we measure it
Before we build, we agree with you in writing:
Match accuracy
Agent results compared line by line with your team’s results on a past period.
Automation rate
The share of items handled without a person.
Exception quality
Whether flagged items are real problems, not noise.
Cycle time
How long the process takes end to end, before and after.
Zero unapproved postings
above the agreed threshold.
What you get
- Production agents for the agreed processes
- Integrations with your ERP, accounting and source systems
- Approval rules and thresholds, documented
- Evaluation set from your historical data, yours to keep
- Exception dashboard
- Runbook and team training
- Source code and configuration in your repository
What we need from you
- A process owner, usually from finance or operations
- Three months of historical inputs and the final reconciled output
- Test access to your ERP or accounting system
- Your matching and approval rules, even if they live only in someone’s head
- Time from the team for a two-week parallel run
Indicative price
₹4 lakh to ₹20 lakh
excluding GST, from pilot to production.
The range depends on the number of processes, systems and rules. The pilot is a fixed price quoted after Discover. 40% advance, the rest on milestones.
Model, API and cloud costs are passed through at cost plus a 10% handling margin when we hold the accounts (no margin if you pay the providers directly). Most back-office agents are cheap to run; we estimate the monthly cost before you sign.
Frequently asked questions
How is this different from RPA?
RPA follows fixed clicks and breaks when screens or formats change. Agents read the content, handle variation and explain their decisions. We still use plain rules wherever they are enough.
Will it post entries into Tally or SAP on its own?
Only if you choose that, and only below a threshold you set. By default, postings wait for approval.
Our CA firm does part of this. Can they use it?
Yes. We can give your CA firm reviewer access to the exception queue.
What if a bank changes its statement format?
The agent usually adapts. If it does not, the failure is caught in the weekly review and fixed under Managed AI.
Is it worth it for a small team?
Sometimes not. If a process takes one person two hours a month, we will tell you to leave it alone.
Name the process your team dreads every month.
A 30-minute call with a founder. We will tell you whether an agent can take it on, and how we would prove it.