Independent research lab · Hyderabad
Market research.An honest record.
moizz is a research lab building measurement tools for Indian equity derivatives. Hypotheses are frozen before new data arrives; results and corrections stay in the record.
- Hypothesis
- Stated in a testable form frozen
- Definition
- Exact rule, universe, data frozen
- Outcome
- What is measured, over what horizon frozen
- Min. effect
- Smallest effect worth detecting frozen
- Hash
- Sealed before the data arrives sealed
- Result
- Recorded, whatever it shows pending
A daily research system, built for measurement.
Every trading day, moizz ingests the National Stock Exchange's bhavcopies (the exchange's daily price files) for the cash and futures & options (F&O) markets, open interest, delivery data and exchange filings. The historical cash archive reaches back to 1994.
Snapshot · 11 October 2026 · these figures describe scope, not performance.
- NSE cash bhavcopy
- F&O bhavcopy
- Open interest
- Delivery data
- NSE & BSE filings
- Cash archive 1994→
Write the question before the answer arrives.
Every registered hypothesis has its definition, outcome, horizon, minimum detectable effect and hash frozen before its data arrives. Results are recorded whatever they show. Corrections are appended, never overwritten.
Define
State the hypothesis, then the exact rule that tests it.
Freeze
Definition, outcome, horizon and minimum effect are sealed with a hash.
Observe
New data arrives after the freeze. Nothing is re-cut to fit it.
Record
The result goes in the register. Corrections are appended, never erased.
Most ideas do not survive. We keep the record anyway.The register exists to show what failed as clearly as what held.
Forward tests run in paper ledgers only: append-only, with hash-locked code. A change to a locked ledger needs a written review by a second AI model and the owner's approval. Hash-locking makes every change traceable; it does not prove the research right, and the record says so when it is wrong.
Check the measurement, not just the result.
Good research fails quietly when the data underneath it is wrong. So the data gets audited too.
PREV_CLOSE is not adjusted on split and bonus ex-dates, so a common corporate-action check never fired.Built with Claude. Claude Code assists development and research checks. Planned next: daily filing analysis and event extraction with the Claude API.
Other work, same discipline.
The same habits, measure first and report what the data shows, applied to company filings and local businesses.
Quarterly results, as filed
Revenue, net profit and EPS for about 1,200 listed Indian companies, straight from their BSE results filings, every quarter since FY19.
Local advertising research
What competitors in a category advertise, with what offers and for how long, read from Meta's public Ad Library.
Business findability audits
A scored rubric for how easily nearby customers can find a local business on maps. An absence of data is never dressed up as a result.
Talk research, or compare notes.
Based in Hyderabad, moizz builds tools for Indian equity-derivatives research and is developing research services for local businesses.
For research conversations and possible collaborations:
Scope. Research and education only. Forward testing is paper-only. moizz gives no trading calls, tips, recommendations or live prices, and nothing on this site is investment advice.