Watchtower
We watch the numbers that matter to you: fuel, rates, rival prices. The moment one crosses a line you set, you hear from us.
Each alert also shows how often it has been a false alarm before, so you know how seriously to take it.
Something big happens in your market. A rival cuts prices, fuel spikes, a new player enters. AstraSight tells you three things: what it does to your profit, in rupees; what the other side is likely to do next; and which response has actually worked before. With proof behind every number.
Our methods are public · Your data stays private
The problem
Pay a consulting firm crores for a study that goes stale in months. Buy data subscriptions that give you numbers but no answers. Or go on gut feel.
Crores for a study that goes stale in months. The team leaves, the deck gets filed, and the watching stops with it.
Cost ▲ · Goes stale · Never scored
They give you numbers but no answers. Nobody tells you what the numbers do to your profit, or what to do next.
Numbers ▲ · Answers ✕ · Never scored
Fast and free. But nobody ever finds out whether the call was right, so nothing gets better next time.
Fast ▲ · Free ▲ · Never scored
And here is the part almost nobody talks about: no consulting firm ever tells you whether its advice actually worked. Nobody keeps score. We built the score-keeping first.
Five service lines
One engine underneath: watching the numbers, scenario planning, war-gaming a fight, pricing the profit impact, and keeping score.
01–05 · One engine underneath
We watch the numbers that matter to you: fuel, rates, rival prices. The moment one crosses a line you set, you hear from us.
Each alert also shows how often it has been a false alarm before, so you know how seriously to take it.
Ask “what if”. What if oil spikes? What if a rival launches? You see the possible outcomes and how likely each one is.
You also see where those odds come from: real past situations, not opinion.
Play out a competitive fight before it happens. See every response open to you, and what happened the last several times companies in your industry tried each one: which worked, which lost money.
Then pick with your eyes open.
Not “this is bad for the sector”. An actual range in rupees: “this event moves your operating profit by ₹X to ₹Y crore, mostly through fuel costs”.
Calculated from your company’s own cost structure, line by line.
Every answer we give is saved, time-stamped and tamper-proof. When the real outcome arrives, we score ourselves, right or wrong, and keep the running tally.
No consulting firm on earth does this.
All five run on the same library of real events, and every answer goes on the same record. A number either shows where it came from, or the engine will not print it.
How it works
We call the method Consequence Intelligence. It follows one event all the way to its consequences: profit, reactions, and your best response. Then it checks, months later, whether we got it right.
Something as simple as “Jio starts a price war”. No forms, no special vocabulary.
It searches a library of real Indian market history for events like yours, and tells you how many it found.
What events like this actually did to companies like yours, in rupees, using your own cost lines.
Each one ranked by how it has really performed before, including the times it lost money.
Early-warning signals that tell you which way things are going, well before the outcome is obvious.
It scores its own answer, right or wrong. The score is kept where it cannot be quietly edited.
Three promises about what we will and won’t say
When we don’t have the history
If we don’t have enough history to answer, we tell you “we don’t know”, in plain words. We do not invent something to fill the gap.
When the answer isn’t what you want
If the move you like has lost money historically, we say so. Even when it’s not what you want to hear.
What the AI is allowed to do
The AI never invents numbers. Every figure is calculated from real, checkable data, or it doesn’t appear at all.
The evidence
No client logos. No testimonials. No accuracy promises. Just what we have built so far, and the checks that keep it honest.
FootnoteEvery figure above can be reproduced inside the product. The test rules were written down before the tests were run.
real competitive battles from Indian markets, studied and measuredacross 14 industries
75
companies where we can turn an event into a rupee profit impact todaybuilt from each company’s own costs
14
We replayed 31 past scenarios with the engine blind to the future. In 30, its recommendation matched what actually worked.the rules were written down first
30/31
automated checks run on the engine, so every claim it can make is testedrun again on every change
2,761
1,058 → 0
We gave the same questions to a leading AI model without our controls: it produced 1,058 figures with no source. Our engine produced zero. Every number is traceable, or it declines to answer.Same questions · Same day · Both answers kept
Time-stamped
Our track record is sealed with public, tamper-evident timestamps. We can’t quietly edit our history, and you can check that yourself.Sealed when written · Anyone can check (OpenTimestamps)
Pricing
We think hidden pricing is part of what’s broken in this industry. For context: that’s less than one strategy manager’s salary, and a fraction of a single consulting study.
All five service lines, all year, run on your company’s own numbers, with your own private record of every answer.
Start free. You pick a scenario, we run it live on your company in front of you, including everything the engine refuses to guess at. If it’s useful, it runs all year. If not, you keep the readout.
First-cohort companies get deeper builds on your own numbers, and priority on new sectors.
Who it's for
Corporates
From ₹500 crore mid-market firms that were never going to pay crores for advice, to listed leaders who want their advisors kept honest.
Consultancies
Run your client workshops on our evidence layer. Your judgment, our receipts, scored on the AstraSight record.
FAQ
Short answers, in the same plain words we use everywhere else on this page.
₹8–25 lakh a year, published right here. That is less than one strategy manager’s salary, and a fraction of a single consulting study. The first readout is free: you pick a scenario, we run it live on your company, and you keep the result either way.
It costs a fraction of a consulting study, and it runs all year instead of stopping when the team leaves and the deck gets filed. The bigger difference: we keep score on our own answers. When the real outcome lands, we mark ourselves right or wrong. Consultants do not.
Data tells you what happened. We tell you what it does to your profit, in rupees, and what to do next. Every response open to you carries the history of how that option performed the last several times companies tried it.
Then we say “we don’t know”, in plain words, rather than guess. Today we have studied and measured 75 real competitive battles across 14 industries, and the library keeps growing.
No. Every figure is calculated from real, checkable data, or it does not appear at all. We put the same questions to a leading AI model without our controls: it produced 1,058 figures with no source. Our engine produced zero.
Send us one scenario you actually care about. We run it live on your company, free: what the event does to your profit in rupees, what your rivals are likely to do, and which responses have worked before. You keep the readout either way.
Contact
Pick something you actually care about. We run it live on your company: what it does to your profit in rupees, what your rivals are likely to do, and which responses have worked before, plus everything the engine won’t guess at. You keep the readout either way.