Cited research
Agents read filings, transcripts, prices and the team's own documents, and return writeups with every number cited to the document it was read from.
Numbers from the data, never a model
v0.1-beta
Scroll for more
[Cursor]
for equity analysts.
Design partner YES Securities
877K events measured
Sitrep is an AI-native operating system for investment teams. Agents research at 100 times the speed of an analyst, cite every number to its source, and every call is scored when the market answers. The team decides.
BSE filings
NSE prices
GDELT
Google News
IIM-A factors
India VIX
[N.01/11]> What it does
Agents do the research.
The team makes the decisions.
Request access
Precedent evidence
For any news item, the similar past events and what each did to the stock, its sector and the market at +1, +3, +5 and +10 days. Cases that went the other way are shown, not filtered.
Shows the cases that disagreeThesis tracking
Theses are versioned, drift is measured, and every call is scored when the window closes and the market answers. The record is the rationale.
Graded against the market[N.02/11]> The corpus
Precedents in seconds.
Not a week of tabs.
Request access
877,067
Measured events
150,045
Filings with full text
+1 / +3 / +5 / +10
Day windows, abnormal terms
43 + GDELT
News feeds, back to 2015
Without Sitrep <> With Sitrep
0100200300400500600700800
[N.03/11]> Core layers
Three layers.
One desk that answers.
Inputs
- Filings
- News
- Prices
- Positions
Contexts
- Company
- Event class
- Sector
- Regime
Outputs
Works with your data
Connect the market data the team already runs on, plus positions, mandates and internal documents. Nothing leaves your tenant.
Built to disagree
It surfaces what the data does not support and the risks a team would otherwise miss.
Graded after the fact
Every output is scored when the market answers. The record cannot be rehearsed or backdated.
[N.04/11]> How it works
Understand the flow.
From news to a graded call.
Request access
// 001
From any source: a filing, a wire story, a broker note, a message from the desk.
A news item comes in
// 002
Same company first, then the same event class across the sector, matched on the text itself, within your lookback.
Agents find the precedents
// 003
Median, range, hit rate stated both ways, and n. The analyst draws the conclusion and seals it with the evidence attached.
The desk reads the evidence
// 004
When the window closes the market answers, the thesis is scored, and the record stands as the rationale.
Reality grades the call
[N.05/11]> Integration
// Add Sitrep to any MCP client: Claude, Cursor, your own agent. { "mcpServers": { "sitrep": { "url": "https://mcp.sitrep.so", "headers": { "Authorization": "Bearer $SITREP_API_KEY" } } } } // 14 tools: precedents, base_rates, thesis, watch, memo, vault ... // Every number the model quotes carries the panel and the document it came from.
from sitrep import Desk desk = Desk(api_key=os.environ["SITREP_API_KEY"]) # One news item in, the precedents out result = await desk.precedents( text="Tata Motors: JLR wholesale volumes down 11% YoY", lookback_years=5, n=20, ) for case in result.cases: print(case.date, case.headline, case.abnormal_5d, case.vs_implied) # Summary across the list: median, IQR, hit rate both ways, n print(result.summary.median_5d, result.summary.hit_rate, result.summary.n)
# Every panel as JSON, with provenance on every number curl -s https://api.sitrep.so/v1/precedents \ -H "Authorization: Bearer $SITREP_API_KEY" \ -d '{ "text": "RBI holds repo at 6.5%, stance unchanged", "subject": "NIFTY", "lookback_years": 8 }' # → { "cases": [...], "summary": { "median_5d": ..., "n": 31 }, # "provenance": { "measured_at": "...", "as_of": "..." } }
Simple to integrate.
Works where you work.
Request access
[N.06/11]> Documentation
Everything documented.
Method, data, results.
Learn more
Everything you need to check us.
The method, the data and the measured results, with the cases that went the other way left in.
[N.07/11]> Invariants
Built to be checked.
Seven things that do not move.
Two clocks
@event_time · @ingestion_time
When it happened and when we learned it are stored separately. A precedent is only what the desk could have known that morning.
Invariant 01 · enforced in code
Read more
Point-in-time membership
Index and sector as they stood on the day. No survivorship.
Verified at the door
@ingestion
Verification status is set when an item is ingested, not when it is read. An unverified item is shown as unverified.
Invariant 03 · enforced in code
Read more
Numbers from the data
@measurement_layer
Every number on a panel comes from the measurement layer and never from a model. CI fails the build if a model output reaches a number field.
Invariant 06 · CI-enforced
Read more
Grades stored per axis
Dated thresholds. A grade can be re-read years later and mean the same thing.
No order capability
@anywhere
Nothing in the system can place, route or size an order. There is no target price and no drafted view. The precedent list is the record.
Invariant 05 · by design
Read more
The analyst concludes
@the_desk
Sitrep retrieves and measures. The reading, the call and the responsibility stay with the person whose name is on the note.
Invariant 07 · the whole product
Read more
[N.08/11]> Pricing
Simple pricing.
One seat, one desk.
- Includes:
- Precedent lookups, every listed company
- Cited research writeups
- Versioned theses, graded
- Email support
- Everything in Analyst, plus:
- Shared coverage, vault and watches
- Dated memos and desk notes
- MCP, Python and REST access
- Adversarial review mode
- Priority support
- Everything in Desk, plus:
- Your positions, mandate and documents
- Your market data connections
- Single tenant, read-only access
- Frozen as-of mode for audit
- A forward-deployed engineer
[N.09/11]> Changelog
We ship fast.
The record is dated.
View all
Every change is a dated commit.
What was measured, what moved, and what did not, in the order it happened.
Sep 14, 2026
Research whitepaper published
The method, the corpus and the measured findings, as a static site at research.sitrep.so. Every claim names its file.
Sep 11, 2026
Filing boilerplate stripped, retrieval re-measured
Covering letters removed from BSE filing bodies. Retrieval improves a little on the hand-checked set, and the number is in the note.
Sep 09, 2026
Macro mode and the similarity floor
460 dated macro events with the Nifty's path after each. Lists drawn across companies now stop at a similarity floor instead of padding.
[N.10/11]> Research
Research notes.
Measured, not argued.
Read all
[N.11/11]> FAQs
We have answers.
Plain ones.
No. Sitrep returns the precedents and the measurement: what similar past events did to the stock, its sector and the market, and the summary across them. The analyst draws the conclusion. There is no drafted view, no target price and no order capability anywhere in the system.
From the measurement layer, never from a model. Reactions are measured in abnormal terms against the sector and the market at +1, +3, +5 and +10 days, and as a multiple of what options implied that day. A model may quote a number, and when it does it cites the panel and the document it read it from.
India first. Every NSE-listed company, with BSE filings, adjusted prices, factor data, F&O implied moves and news back to 2015, plus 460 dated macro events for the index itself. Any market with an exchange feed comes next.
Yes. Sitrep ingests the team's coverage, positions, mandate and internal documents, and connects to the market data you already run on. Access is read-only, revocable, and single tenant on the Enterprise plan. Nothing is used to train a model.
Every thesis is sealed with its evidence and a window. When the window closes, the market answers and the thesis is scored per axis against dated thresholds. Sitrep's own panels are graded the same way. Nothing can be rehearsed or backdated.