NSE closed·  :  :   IST
sandbox · v0.1
VajraQuant

Trading operating system · India

Describe a strategy.
Read the code.
Watch it trade.

A single workbench for research, strategy, testing, execution and tax — built for the Indian market, and for the person who trades their own money. Every strategy you run is code you can read.

join the waitlist →see the surfaces

signature backtestorb-nifty-15m · 5y · seed 42 · reproducible

cagr
+38.2%
sharpe
1.84
max dd
-9.4%
win rate
58%
trades / yr
146

The terminal

mock data · 09:41 IST · sandbox
for the self-directed trader

Describe it. Read it. Test it. Run it.

Write a strategy in plain English and get Python you can actually read. Backtest it with real costs and an overfitting verdict that argues back. Run it on paper against a live-shaped execution path, with risk limits that stop the account, not just the strategy.

  • plain-english to readable python
  • costs and slippage included by default
  • paper execution that mirrors live
  • kill switch always in reach
see the features →
for HNIs and family offices

One answer to “what do we own, today.”

Multiple entities, multiple custodians, one consolidated view. Every position, order and cash movement hangs off a legal owner rather than a login — built that way from the first migration, because retrofitting it means rewriting the data layer.

  • individual, HUF, trust, LLP, company
  • per-entity books and per-entity tax
  • entity isolation enforced in the database
  • a CA seat scoped to one entity
for family offices →

What it does

Everything the loop needs, in one product

strategy

Plain English to code

Describe the idea; read the Python it becomes. Import what you already have from Pine Script, AFL, or a spreadsheet.

backtest

Tests that argue back

Real slippage and the full Indian charge stack, with an overfitting verdict above the fold instead of buried in a stats table.

execution

Paper that mirrors live

Full order lifecycle, resting orders, partial fills, rejections — writing into the same tables live execution will use.

risk

Limits on the account

Per-strategy checks plus a book-level drawdown governor that halts everything, including the strategy that was winning.

tax

A file your CA can work from

FIFO lots, §112A grandfathering, ICAI turnover, Schedule 112A, AIS reconciliation, and a scoped read-only CA seat.

trust

Evidence, not assurances

Tenant isolation in the database, an audit trail on every change, and encrypted broker credentials read only through an audited reveal.

See every feature, and what is still ahead →

Five surfaces, one loop

01

P1 · terminal

data, news, charts, screeners — the surface you watch.

read →02

P2 · strategy coder

plain english compiles to code you can read.

read →03

P3 · orchestration

deploy, monitor, kill — the kill switch is two taps away.

read →04

P4 · capital allocator

position sizing, risk budgets, drawdown governors.

read →05

P5 · strathub

publish and fork strategy code — never signals.

read →

The thesis, stated plainly

the loop

data → idea → code → backtest → deploy → attribute → publish, in one product. Bloomberg gives data and news but no strategy lifecycle. QuantConnect gives an engine but is broker-restricted. TradingView gives charts but no real execution.

white-box, always

every strategy you run is code you can read. no black-box signals. this is why we don't need SEBI research-analyst registration for the core product — and why you're never trading logic you can't audit.

overfitting made visible

in-sample vs out-of-sample by default. a concentration warning when three days of March produced the whole year's returns. a plain-language verdict at the top of every backtest — not buried in a stats table.

the kill switch is sacred

per-strategy, per-account, global. always visible in app chrome. two taps maximum from any mobile screen.

waitlist

start when we're ready.

we'll email once when there's something to sign up for. no drip, no digest.

no drip, no digest. one email when there's a real product to sign up for.