AI analytics platform · A GeXOn AI product
Know why sales, cash and customers changed. Ask, and check the answer.
Ask in plain words. SolmiriX answers from the systems you already run, read-only, and shows the figures behind every answer. Forecasts, alerts and a weekly report come with it.
Why did revenue decline?
Answer Revenue fell 11.3% in March, from Rs 40.7 L to Rs 36.1 L. Koshi accounts for 80% of the fall, and three Koshi wholesale accounts that stopped ordering after the 4 March price change explain Rs 3.4 L of it.
| Region | Jan | Feb | Mar | Apr | May | Jun |
|---|---|---|---|---|---|---|
| Bagmati | 16.2 | 16.8 | 16.5 | 17.1 | 17.6 | 18.2 |
| Koshi | 9.4 | 9.8 | 6.1 | 6.6 | 7.9 | 8.8 |
| Gandaki | 6.1 | 6.0 | 5.8 | 6.2 | 6.4 | 6.6 |
| Lumbini | 7.8 | 8.1 | 7.7 | 8.3 | 8.8 | 9.2 |
From Sales ledger · CRM · Price list
01 The problem
When one person knows the spreadsheet, every answer waits for them.
The data already exists: in the CRM, the invoicing tool, the support desk and a few shared spreadsheets. Getting an answer out of it usually means asking the one person who knows how it fits together, and waiting.
- What happens today
- The same customer is listed three ways across three systems, and someone reconciles them by hand at month-end.
- Why it matters
- Reports take days, only one or two people can run the numbers, and when they are on leave the questions wait.
- What changes
- One cleaned, connected layer, and anyone on your team can ask it a question in plain language.
- What improves
- Everyday questions stop queueing for a specialist, month-end needs less reconciling by hand, and problems show up while you can still act on them.
02 Module · Ask AI
Ask in plain language. Get the figure, the reason, and where it came from.
The question is one sentence. The answer says what happened, why it might have happened and what to check next, and every number opens to the rows it came from.
Step 1 · Ask
Why did revenue fall in March?
11:02 AMAn ordinary sentence — no query syntax, no field names. Nepali works the same way, and the answer comes back in the language you asked in.
Step 2 · Plan
- MeasureRevenue, monthly
- PeriodJan – Mar 2026
- Break down byRegion, then customer
- CompareFebruary vs March
It says what it thinks you meant before it looks anything up — so a misunderstanding costs you one click, not one meeting.
Step 3 · Look
- ✓Sales ledger
- ✓CRM
- ✓Price list
- Permission check
Named sources, and only what you are allowed to see.
Step 4 · Answer
MeasuredRevenue was Rs 40.7 L in February and Rs 36.1 L in March.
CalculatedA fall of 11.3%. Koshi accounts for 80% of it; three Koshi wholesale accounts explain Rs 3.4 L.
AI · possible reasonTheir orders stop right after the 4 March price change.
AI · check nextAsk the Koshi sales lead whether the new prices reached these accounts.
11:02 AMके अप्रिलमा कोशी फर्कियो?
11:03 AMआंशिक रूपमा। कोशीको राजस्व मार्चमा रु ६.१ लाखबाट अप्रिलमा रु ६.६ लाख पुग्यो, तर फेब्रुअरीको रु ९.८ लाखभन्दा अझै कम छ।
11:03 AMAsked in English, answered in English. Asked in Nepali, answered in Nepali — the follow-up above switched language and the reply followed it.
Step 5 · Trace
- The 11.3%Open the order rows behind it ✓
- ExportExcel, PDF or API ✓
- Follow up"Which customers stopped ordering?"
- ShareSend the answer with its sources
An answer you cannot check is an answer you should not act on.
- Measured read from your systems
- Calculated arithmetic on those rows
- Forecast a range, not a promise
- AI a reason or next step, for a person to judge
-
Ask
A question in ordinary words, in English or Nepali. No query language, no field names, no waiting for the one person who knows how the spreadsheet is put together to come back from leave.
-
Plan
Before it looks anything up, it says what it thinks you are asking for — which measures, which period, which breakdown. You get to catch a misunderstanding here, at the cheap end, instead of arguing with a wrong number later.
-
Look
It reads from the unified layer, and only the parts you are permitted to see. Which sources it touched is shown, so an answer is never a black box — and someone who cannot see all regions does not get an answer that silently includes them.
-
Answer
What happened, why it might have happened, and what to check next, with each line marked: measured from your systems, calculated from those rows, or suggested by the AI for a person to judge. You can read it out in a meeting without translating it first.
-
Trace
Every figure links back to the rows behind it. Export it, or ask the obvious follow-up and keep going. An answer you cannot check is an answer you should not act on.
- Questions in plain English or Nepali
- Sources named, permissions respected
- Measured, calculated and suggested, kept apart
- No query language for everyday questions
03 Module · Data unification
Five inconsistent exports. One structured layer.
Count how many places one customer lives. SolmiriX matches and merges those records automatically — and remembers which source every field came from.
Sources 5
- CRM
- Invoicing
- Inventory
- Spreadsheets
- Support desk
Step 1 · Connect
- AccessRead-only ✓
- Migration neededNone ✓
- WorkspaceIsolated ✓
- Total rows15,738
Nothing is moved or replaced. Your systems stay exactly where they are.
Step 2 · Profile
- customercust_name
- customerCustomer Name
- customeracct_nm
- ⚠region
Three names, one thing. The gaps get counted before anyone builds a number on top of them.
Step 3 · Clean
- 03/04/26 · 2026-04-03One format ✓
- "NPR 64,200"64200.00 ✓
- arun wholesale · ARUN WHOLESALE PVTArun Wholesale ✓
- 214 junk rowsQuarantined ⚠
Quarantined, not deleted — you can open the pile and check what it decided.
Step 4 · Merge
- Arun Wholesale3 records → 1
- Emailfrom Invoicing ✓
- Regionfrom CRM ✓
- Phone conflictCRM wins · kept in history
Every field remembers where it came from. You can always ask why a figure says what it says.
Step 5 · Serve
- Ask AIPlain-language questions ✓
- DashboardsStandard reporting ✓
- ForecastingPredictive modules ✓
- ExportExcel, PDF or API
One layer, refreshed on a schedule — so it does not go stale the week after it is built.
-
Connect
Read-only connections to the systems you already run — CRM, invoicing, inventory, the shared spreadsheets, your support desk. Nothing is migrated and nothing is replaced. Your tools carry on being the systems your team uses.
-
Profile
It works out what each column actually is.
cust_name,Customer Nameandacct_nmare recognised as the same thing, types are inferred, and the empty and malformed values get counted before anyone builds a number on top of them. -
Clean
Formats are normalised so the same fact is written the same way everywhere — dates, currency, casing, phone formats. Rows that are genuinely junk are quarantined where you can inspect them rather than silently dropped.
-
Merge
Records that describe the same customer collapse into one, with a lineage trail showing which source each field came from. When two sources disagree, you decide which one wins — and the losing value is kept, not erased.
-
Serve
The result is one structured layer that everything else reads from: Ask AI, your dashboards, exports, and the forecasting modules further down this page. It refreshes on a schedule, so it does not go stale the week after it is built.
- Auto-schema & field mapping
- Deduplication & cleaning rules you control
- Connectors for the tools you already run
- One structured source of truth, with lineage
How it fits around the systems you run
Sources
What you already run
- Excel & Sheets
- CRM
- ERP
- Accounting
- Databases
- Support desk
Connect & clean
Profile, fix, deduplicate
- Read-only connectors
- Schema mapping
- Quality checks
- Quarantine, not delete
Unified layer
One structured model of your business
- Isolated per organisation
- Field-level lineage
- Permission-aware
Six modules · one platform
What you do with it
- Data unification
- Ask AI
- Forecasting
- Anomaly detection
- Recommendations
- Auto-reports
- Your systems
- Stay where they are. Connections are read-only; nothing is migrated or replaced.
- Your data
- Lives in a workspace isolated to your organisation, with access logged.
- Every figure
- Traces back to the source rows it came from — and which system won any disagreement.
- Every answer
- Respects who is asking. People only see figures from sources they may already read.
04 Six modules
Two foundations. Four modules that look ahead.
Data unification and Ask AI are the foundation: one clean layer, and questions answered from it. Four predictive modules read from that same layer.
-
Foundation 1
Data unification
Scattered exports become one structured layer, with the source of every field kept.
-
Foundation 2
Ask AI
Plain-language questions, answered with the figures and rows behind them.
Predictive module 1 · Forecasting
See next quarter’s revenue as a range, while there is still time to act.
Once your data is unified, statistical and machine-learning models project demand, revenue and cash flow from your own history and pipeline, instead of everyone relying on last month’s numbers.
Proof detailEvery forecast is a range, drawn apart from the history it came from, so nobody mistakes a projection for a result. Models retrain as your history grows.
Forecast Q3 revenue: Rs 1.33 Cr, range Rs 1.26–1.39 Cr
Show the numbers
| Month | Value | Range |
|---|---|---|
| Jan | 39.5 | |
| Feb | 40.7 | |
| Mar | 36.1 | |
| Apr | 38.2 | |
| May | 40.7 | |
| Jun | 42.8 | |
| Jul (forecast) | 43.5 | 42.0–45.0 |
| Aug (forecast) | 44.2 | 42.1–46.3 |
| Sep (forecast) | 45.0 | 42.3–47.7 |
Predictive module 2 · Anomaly alerts
Catch a drop the day it happens, not at month-end.
SolmiriX watches your key metrics against their normal range and sends an alert to the person who owns the metric the moment one breaks pattern.
Proof detailHere, Koshi’s daily orders fall outside their expected range on 5 March, the day after the price change, and the regional sales lead hears about it that morning, weeks before the monthly report would have shown it.
Calculated Koshi orders −31% vs expected since 5 March
Predictive module 3 · Recommendations
Get a suggested next step, with its reasoning, for a person to approve.
One step past forecasts and alerts: SolmiriX suggests a specific action and shows the evidence it used. Nothing happens until the person who owns the decision approves it.
Proof detailEvery suggestion is logged with who approved, changed or dismissed it, and what happened next.
Predictive module 4 · Auto-reports
The monthly report, written in plain language and sent on time.
The narrative report someone assembles by hand every month: written from the same layer, in your template and tone, and delivered on a schedule.
Proof detailEvery figure in the report links to the rows behind it, and it goes out as PDF or Excel, or through an API.
Revenue rose 5.2% on May to Rs 42.8 L, led by Bagmati at Rs 18.2 L.
- Koshi reached Rs 8.8 L, up from Rs 6.1 L in March but still below February’s Rs 9.8 L.
- The three Koshi accounts that stopped ordering in March still hold the oldest overdue invoices.
- Forecast July: Rs 42.0–45.0 L.
05 Which route fits
Answers without a reporting project.
Most owners wait for someone to build a report before they can see a number. SolmiriX answers from the systems you run today. If you already pay for Power BI or Tableau, keep it: SolmiriX answers the questions your dashboards weren’t built for.
| Category | Both | Reports on your BI tool | |
|---|---|---|---|
| Fits when | People ask new questions every week, and the answers sit across several systems. | You check the same numbers every week, and new questions keep coming in between. | You track the same numbers every week, or need fixed board or management reports. |
| You get | Answers about sales, cash and customers, with the figures behind them. Forecasts, alerts and scheduled reports come with it. | Dashboards for the numbers you check weekly, and SolmiriX for the questions in between. | Pipelines, one set of metric definitions and dashboards on Power BI, Tableau, Looker Studio or Excel. |
| Who runs it | We host it, in a workspace isolated to your organisation. | Each part as in the columns either side. | You, on your own licences, with documentation and a handover. |
| First answer | As soon as your sources are connected and cleaned: days for a straightforward setup, longer for tangled data. We tell you which, in writing, after looking. | Scoped together. | The first dashboard comes early in the build; the scope states the timeline. |
| Priced as | In writing, after the walkthrough, for your sources and users. | Scoped together. | A fixed price in a written scope. |
SolmiriX does not replace an analyst’s judgement: complex modelling and one-off investigations still benefit from one. What changes is that everyday questions don’t wait for a report to be built. Reports on your BI tool
06 Security and data control
Your data stays yours, and you can see what is done with it.
What is in place today, and what is not yet, stated plainly. The full detail is on our Security page
In place today
- Read-only connections to your systems; nothing is migrated
- A workspace isolated to your organisation, never used to train shared models
- Answers that respect who is asking: people only see figures from sources they may already read
- Access to your data limited to the engineers working on it, and logged
- Bad rows quarantined, not deleted, so you can check what was decided
- Recommendations wait for a person’s approval; nothing acts on its own
- Encrypted in transit (TLS 1.2 or later)
- Full export on request, and connected data deleted within 30 days of offboarding
Not in place yet
- A third-party penetration test
- SOC 2, ISO 27001 or an equivalent certification
- A published sub-processor list (we send the current list on request)
- A published uptime SLA
If your security review needs “not applicable” or “in progress” against any of these, we fill it in honestly rather than leave it blank.
07 Who it is for
The people who need the numbers ask for them directly.
No hire needed for everyday questions. Each role asks in its own words and gets an answer it can check.
Owner or managing director
“Why did revenue fall in March?”
Decisions made on this month’s numbers, not last quarter’s.
Finance lead
“Which invoices are overdue, and with whom?”
Month-end with less reconciling of exports by hand.
Operations manager
“Which region’s orders dropped this week?”
Problems caught the day they happen.
Sales lead
“Which customers have stopped ordering?”
Lost customers noticed while they can still be won back.
Where scattered data costs most
- Retail & distribution
- Manufacturing
- Logistics
- Financial services
- Healthcare
- Education
- Hospitality
08 How it starts
The software is ready. Your data decides the timeline.
All six modules are in production use today. What varies between businesses is the data: how many systems you have, how consistent they are, and how much cleaning the first pass needs. We tell you which yours is after we have looked.
-
Free · 30–45 minutes
Walkthrough
We look at the systems you run and show SolmiriX answering the kind of questions you ask.
-
In writing
Data assessment
We check your sources and tell you how long the first connection takes, what it needs from your side, and what it costs.
-
Days to weeks, depending on your data
Connect and ask
Read-only connections, cleaning rules agreed with you, and then your team starts asking.
If the platform is not the right answer, we can build reports on your BI tool, or our automation services can take the repetitive work off your team directly.
09 Questions
Questions about SolmiriX.
Do we need to hire anyone?
Usually not for everyday questions and reports: the people who need the numbers ask SolmiriX directly. Someone on your side owns the data rules, such as which system wins a conflict, and that is usually a finance or operations lead, not a new hire. For complex modelling or one-off investigations, an analyst still adds value.
Where does our data live?
Your systems stay where they are; SolmiriX reads from them read-only. The unified layer lives in a workspace isolated to your organisation, and we name the hosting provider and region in writing before you connect anything.
Can we leave with our data?
Yes. Your source systems never move, so they stay yours. On request we export the unified data, and we delete the connected data within 30 days of offboarding, except records we are legally required to keep.
Do we have to move our data into SolmiriX?
No migration. SolmiriX connects to your systems read-only and builds the unified layer in an isolated workspace for your organisation. Your systems stay exactly where they are.
Who can see what?
Answers respect permissions. A question only ever returns figures from sources the person asking is allowed to see.
How long does setup take?
It depends on your data, not the software. A straightforward setup connects in days; a tangled one takes longer. After looking at your systems we tell you which yours is, in writing.
Does it work in Nepali?
Yes. Ask in English or Nepali and the answer comes back in the language you asked in.
What if SolmiriX is not the right fit?
Then we will say so. If you need fixed reports, we can build them on your BI tool; if the problem is repetitive work, our automation services take it off your team, without the platform.
See SolmiriX answer questions about your own data.
A walkthrough on a sample of your own exports shows you more than any demo on ours. Tell us what systems you run and what you wish you could ask them.