P
PRAMANA
RESEARCH
STOP RENTING STRATEGIES YOU CANNOT CHECK

Build a professional
trading system.
With AI. Without code.

Long-term, swing and short-term strategies — built by you, tested on 30 years of Indian market data, and explained number by number.
🤖No coding knowledge needed — AI writes it, you decide it
📊You will understand every number — worst fall, Sharpe, recovery, real costs and tax
🌐Sessions and demos are conducted in English
No tips. No signals. No profit promises. Just the method.
S
Santhosh
Self-taught developer  |  Systematic market researcher  |  Built Pramana alone

What you learn to do

5 Checks That Separate a Real Result
From a Flattering One

Spot Survivorship Bias

Spot Survivorship Bias

Find the 31% of the market your data quietly deleted — and see what it does to a return.

Catch Corrupted Prices

Catch Corrupted Prices

One sort that exposes broken corporate-action maths before it flatters your result.

Charge Yourself Like a Real Broker

Charge Yourself Like a Real Broker

Full costs and real Indian capital-gains tax — the switch that moved profitable set-ups from 4 to 17.

Prove There Is No Look-Ahead

Prove There Is No Look-Ahead

A leak test you verify by planting a bug yourself, then watching the test catch it.

Measure the Luck

Measure the Luck

How much of your best result is chance, not edge — and the maths that separates them.

“A backtest is not a promise of what will happen. It is a record of what you would have had to sit through — and most people never check whether they could have.”

Where this leads

Build your own trading system.
Then understand every number it gives you.

Most results lie. This is where you build the thing that does not — long-term, swing and short-term strategies, from scratch, on 30 years of Indian market data. Using AI, with no coding knowledge required.

1

You build it, so you understand it

Not a strategy handed to you. A system you assembled yourself — which means when it stops working, you will know where to look instead of waiting for someone to tell you.

2

Every number, explained properly

Not just the return. What the worst fall really means, whether a Sharpe ratio is trustworthy, how long recovery took, and why your money-weighted return differs from the headline.

3

See how it behaves when things go wrong

Run the same strategy through 2008, 2020 and the quiet years in between. A system you have only seen in a bull market is a system you have not seen.

4

Test it properly, not once

Hundreds of parameter combinations, walk-forward testing, and the correction that tells you how much of your best result was luck rather than edge.

THE NUMBERS YOU WILL LEARN TO READ

Not just the return — the numbers that tell you whether you could actually have lived through it.

Worst fall Time under water Sharpe ratio Recovery time CAGR vs XIRR Win rate vs payoff Cost & tax drag Luck-corrected score

THE CONTROLS YOU WILL LEARN TO SET

Market filter — what to do in a falling market Position sizing Rebalance schedule Universe & liquidity limits Real broker costs & slippage Indian capital-gains tax Dividends Buy limits & dead-stock rules Leverage & margin interest Monthly SIP & withdrawals Stops, targets & time limits
🤖

No coding knowledge needed — AI writes it, you decide it

You will not be learning Python syntax. You will be learning what to ask for, what to check in what comes back, and when to refuse the answer. AI does the typing; the judgment stays yours, because that is the part that is actually worth having.

FREE · ONE TO ONE

See it running on my screen, and ask me anything

Register below and I will call you back to fix a time. Then, one to one, I will walk you through the platform live, run a strategy end to end, and answer whatever you want to ask — including whether this is wrong for you.

30 minutes, screen shared, one to one — never a group session
See exactly what you would be building, and what it will take from you
Bring a backtest you already have — we will check it together
Conducted in English, at a time that suits you
No obligation, and no stock recommendations of any kind
Request My Free 1:1 Demo →
☎ I call you back to schedule · limited slots each week

Is this for you

Everybody backtests.
Almost nobody checks the backtest.

Self-taught algo traders

You have built a backtest. You are not certain you could defend it to someone who asked hard questions.

Swing & positional traders

You follow rules, but you have never tested them on data that included the companies that died.

</>

Python learners

You can write the code. Nobody has told you what to check before you believe the output.

Finance professionals

See where a research process actually earns its credibility — and where most of them quietly lose it.

First-time systematic investors

Start with a method you can verify instead of a tip you have to trust.

Anyone sold a “90% accuracy” claim

Learn to check the claim yourself, in minutes, without taking anybody's word for it.

What you will be building

Build your own system.
Understand every part of it.

Anyone can hand you a strategy. Here you build the thing you can actually defend — long-term, swing and short-term strategies, from scratch, on 30 years of Indian market data.

Not a strategy handed to you to copy. A system you built yourself, whose behaviour you can explain to anyone who asks — including yourself, at the worst possible moment.

Three horizons, one honest engine

Long term

months to years

Rank the whole market on a schedule, hold the leaders, and let the reshuffle do the risk management.

Swing

days to weeks

Wait for one specific setup, take it, and exit on a stated reason — never on a calendar.

Short term

tight holds, end-of-day driven

Shorter holding periods, where costs and slippage decide whether an edge survives contact with reality.

The numbers you learn to read — and read together

Drawdownthe worst fall

How far your money dropped from its highest point. The number that decides whether you could actually have held on.

Sharpe ratioreturn per unit of risk

Two strategies both returning 15% are not equal if one lurched to get there. This is how you compare them fairly.

Recoveryhow long the pain lasted

Not just how deep the fall went, but how many months you sat underwater waiting to get back to even.

Profit factorrupees won per rupee lost

A blunt honesty check — and you learn to recompute it with the ten best trades removed.

Exposurehow often you were actually invested

A strategy that sat in cash for six years is not the same as one that was at risk every single day.

Trades & win rateread together, never alone

A 70% win rate means nothing until you know the size of an average win against an average loss.

The controls you learn to turn, and what each one really changes

Market filterFive ways to react to a falling market — the single most valuable rule found in the whole project.
Position sizingEqual weight, smaller for jumpy stocks, or larger for the highest-ranked.
Costs & slippageThree real brokers, and impact that grows with the size of your order.
Real Indian taxNetted by financial year, losses carried forward, the LTCG exemption applied.
Buy limitsNever buy more of a stock in a day than it actually traded that day.
Dead-stock rulesWhat a delisted holding is worth — stated openly instead of quietly assumed away.
LeverageModelled as a real loan: daily interest, margin calls, and the date the account was wiped out.
Money in & outMonthly contributions and withdrawals — and why CAGR gets switched off when money moves.
Walk-forwardRe-pick using only past information, then test on the future you had not seen.
Rolling windowsSteadiness across time, instead of one flattering number from one lucky window.
Monte CarloThousands of re-runs with the fills shaken — seeded, so the claim is repeatable.
Market phasesThe same strategy scored separately through bulls, crashes and years of going nowhere.

“How does it behave when the market changes?”

That is the question the whole cohort is built around. You will run the same strategy through bull markets, through crashes, and through years of going nowhere. You will test it on data it has never seen, re-run it thousands of times with the fills shaken, and watch what moves when you change one control at a time. By the end you will not be hoping a system works. You will know how it behaves, what it costs you, where it breaks — and you will be making your own decisions about it.

Request My Free 1:1 Demo →
☎ Register and I call you back to fix a time

What this is not

You will not get stock recommendations, buy or sell calls, or a strategy to copy. I am not SEBI-registered and I do not give advice. What you get is a way of thinking you can apply to any strategy — yours or anyone else’s — and a one-page checklist you will keep using long after the session ends.

Your Instructor

Meet Santhosh

Self-taught developer  |  Systematic market researcher  |  Builder of Pramana

I taught myself to build this, and I built it alone — a research platform that runs strategies across three decades of whole-market NSE data, keeping every company that later died. Everything I know about testing a strategy honestly, I learned by getting it wrong here first.

Everything I teach came out of that build, including the parts I would rather not show you. One of my own backtests once bought 6×1034 shares of a company at a fraction of a paisa and reported a profit. It did not crash. It looked completely normal. That single afternoon is why the platform now has guards, audits and a leak test that must catch a bug I plant on purpose before I trust it.

The most useful thing this platform does is tell me when a result is not as good as it looks — and it has told me that more often than I would like. A research process that can only ever agree with you is not a research process. It is an advert with charts.

Years of NSE data30
Stocks kept, dead ones included11,339
Companies that stopped trading3,505
Automated tests passing264
Configurations logged & corrected192
Stock tips ever given0
Request My Free 1:1 Demo →
☎ I call you back to fix a time
Request My Free 1:1 Demo FREE
☎ Register and I call you back to fix a time