Research, not signals.
A research brief in your inbox before NSE opens, followed by a public audit of yesterday's calls.
Or read the philosophy first — six minutes, no signup.
Live preview · redacted public view
The record so far
A share of the early figures is drawn from retrospective replay of historical sessions, marked as such in every audit. Live-collected numbers grow each session.
since launch, across equity, options, index
every claim resolved, hit or miss, on the next session
rolling 30 trading days, touch-watch head
no claim is published without an outcome the next day
Discipline 01
The Daily Brief gives you market structure, sector pressure, touch watch, reaction context, and avoidance flags before the session begins. It is built for a trader who wants better preparation, not a command to follow.
Public samples redact exact levels and instruments where needed. Subscriber briefs carry the full view inside the portal, while public pages keep the product understandable without exposing the research machinery.
The decision to act on any of our research remains entirely with you, the reader. Read why this matters
We grade ourselves out loud, every morning before the bell.
Discipline 02
Tomorrow’s brief opens by auditing today’s. Per-confidence-bucket hit rate. Calibration error per prediction type. Drift flags surfaced before they degrade.
When the audit is drawn from a retrospective replay of historical bundles rather than live-collected outcomes, we say so explicitly, and the brief carries a disclosure line so you never confuse the two.
No other research desk in this space audits itself publicly. See how the audit works
Discipline 03
A 70% probability we publish should resolve in our favour roughly 70 times out of 100. If it doesn’t, the gap shows up on the dashboard as calibration error, and we flag the head as drifting until it returns to within tolerance.
The whole engine is built around this contract. Every probability you read was calibrated against a hold-out window the engine never saw during fitting.
Daily mean predicted probability minus actual hit rate, last 90 trading days. Closer to zero is better.
From the desk
Of Garvit Katyal · founder · Crux Research
When I started building this, I wasn’t trying to invent another signal service. India already has thousands of those. I was trying to answer a smaller, more honest question: when a brief says “70% likely to touch,” does it actually happen 70 times out of 100?
Most desks never check. We do, every morning. The dashboard you just scrolled past is the same dashboard I look at before I write the next brief. When a head drifts, you see it before I do.
If that kind of accountability is what you want from your morning read, the sample brief is the place to start. If it isn’t, no hard feelings — close the tab.
Garvit
Bangalore · India
published before the bell
How a brief comes together
We don’t hide the pipeline. We hide the technique. The stages below run in this order every NSE session — the same shape, the same gates, the same audit at the end. Subscribers see the output; the engine’s internals stay inside the moat.
Ingest
Overnight feeds and prior-session prints arrive.
Weave
Sector regime, structural levels, and reaction context are stitched.
Calibrate
Every probability is checked against a hold-out window.
Audit
Yesterday's brief is graded against actual outcomes.
Publish
Brief lands in the inbox before 08:30 IST.
Stage labels are the engine’s published pipeline names. Specific models, features, and tuning parameters stay inside the subscriber portal where they belong.
Built for
We don’t segment by tier or asset class. We segment by how you use the brief. If you recognise yourself in one of these, the product is built for you.
01 · The independent trader
You're past the tipster phase. You want pre-market context that respects your judgement, not a buy call you have to second-guess. The brief tells you where the structural levels are, where the regime is leaning, and where to stand aside — then it gets out of the way.
02 · The desk analyst
You want a sober second opinion before you walk into the room. Calibrated probabilities, redacted sector regime, and an honest audit of yesterday's calls give you something to anchor or argue against — not a black-box signal you can't explain.
03 · The fund PM
You don't need another data feed. You need a research partner whose calibration error you can verify in public, whose drift is flagged before the loss is taken, and whose discipline you can show to your IC.
The reader makes the trade. The brief makes the case.
How we publish
We sell context — sector regime, structural levels, reaction quality, avoidance flags. We don't sell instructions. The decision to act remains with the reader.
Every claim carries a number. A 62% touch probability is the system's calibrated estimate of how often historical analogues did, not a prediction it will.
70% should resolve 70 times out of 100. When it doesn't, the gap shows up as calibration error and the head is flagged drifting until it returns to tolerance.
Every brief opens by auditing yesterday's. Hit rate per confidence bucket, calibration error per prediction type, drift flags before they degrade the next read.
Pricing
Google sign-in starts a five-day preview automatically. Paid plans run through Razorpay and keep the exact research archive inside the subscriber portal.
Equity, options, and index research briefs with archive access. Live intraday updates are not included.
Core Research plus live update delivery, full archive, calibration dashboard, and priority support.
One-off research or infrastructure audit. Starts with a base scope and expands by infrastructure size and data complexity.
Common questions
If your question isn’t here, the contact page reaches the founder directly.
Neither. The Daily Brief is a research artifact — sector regime, structural levels in play, reaction quality, avoidance flags. Every claim carries a calibrated probability. The decision to act remains with the reader.
No. We are a research publisher, not a registered investment advisor. The brief is research context only, not investment advice. The disclosures page is explicit about this and every page footer carries the short form.
When the brief publishes a 70% probability, that estimate is checked against the actual outcome rate on a holdout window the model never saw during fitting. The dashboard publishes the gap (calibration error) per confidence bucket. If a head drifts out of tolerance, we flag it until it recovers.
Briefs are published by 08:30 IST on NSE trading days. Email plus the subscriber portal. The portal carries the full archive.
Yes — the sample brief is a real published artifact, redacted for public view. Subscriber-only levels are masked but the shape, sections, and language are intact. Google sign-in also starts an automatic five-day preview.
Nothing. Your account stays. The preview is a single five-day window — when it expires, the portal goes back to the public marketing surface. No card charged, no follow-up sequence.
Subscriber briefs publish exact structural levels and reaction context. They are not entry/exit instructions — they are the context against which the reader's own decision is made. The brief never tells the reader what size, when to enter, or when to exit.
Monthly subscriptions cancel any time from the portal. Annual subscriptions are refundable pro rata within the first fourteen days after billing.
Six minutes · no signup · no card
The sample brief is a real published artifact, redacted for public view. No follow-up sequence, no aggressive retargeting. Read it, close the tab, come back when you want to.