For AI builders

Your model reads it.
No preprocessing needed.

Konseki returns structured JSON with outcome distributions, reliability scores, and pre-written commentary. Drop it into a prompt — your model understands it natively and reasons from real historical evidence.

Get free API key View API docs →

Your prompt

// system prompt
You are a trading assistant. Answer using only the provided market context data. Be concise and cite specific numbers.

// user message
What does AAPL's current setup look like? What usually happened historically in the different forward windows after this setup? Is it a good timing to get into the market?

// context injected automatically
{ Konseki JSON → see below }

Konseki JSON injected

Model response

Click "See example response" to see
Claude reason from the Konseki data

How it works

Call the API once per symbol

One request to GET /v1/analysis/{symbol} returns everything: outcome distributions, match quality, MAE/MFE, reliability tags, and pre-written commentary. No secondary calls. No joins.

Inject the JSON directly into your prompt

No transformation required. The response is already structured the way a model reasons — named fields, natural-language commentary, numeric evidence. Paste it in as-is.

The model cites real evidence

Your assistant can now say "historically, 60% of similar setups were positive over 5 days, with a median move of +3.4%" — because the evidence is right there in context, not hallucinated.

Why this matters for AI assistants

Models don't have market memory

LLMs have broad financial knowledge but no access to what this specific stock looked like the last 15 days, or what happened next in similar historical setups. Konseki provides exactly that — per-symbol, per-day, computed fresh every close.

The commentary field is prompt-ready

Every response includes a commentary object with pre-written fields: headline, summary, risk, takeaway. Your model can quote them, rephrase them, or use them as reasoning scaffolding — they're designed for LLM consumption.

Grounded answers, not hallucinated statistics

When your assistant says "similar setups had a worst historical outcome of −7.6%", that number came from the JSON you provided. It's verifiable, sourced, and specific — not a statistical generalization your model invented from training data.

No preprocessing pipeline to build

Financial data typically requires normalization, feature engineering, and interpretation before a model can use it. Konseki handles all of that upstream. You get a JSON response that's ready to inject — no pipeline, no transformation layer, no maintenance.

Key fields your model will use

commentary.summary

Pre-written natural language summary of the historical evidence. Drop it straight into context.

returns.positive_return_rate

What fraction of historical analogs finished positive over this forward window. A concrete, citable number.

tags.direction / tags.risk

Categorical signals your model can use for tone calibration: bullish_moderate, high_tail, etc.

returns.max_adverse_excursion

Worst intraday drawdown across historical cases — lets your assistant give honest downside context.

What builders are using it for

Chat assistant

Stock Q&A with historical grounding

User asks "how has AAPL behaved after setups like this?" — your assistant fetches the Konseki JSON and responds with actual historical outcome distributions, not general market knowledge.

Research copilot

Pre-trade context generation

Before a user reviews a position, inject the Konseki JSON into a system prompt. The model surfaces relevant risks, positive rates, and match quality without the user having to ask.

Alerts & digests

Automated morning briefings

Fetch Konseki data for a watchlist every morning, pass each symbol's JSON to a model, and generate a structured briefing: what setups are elevated risk, what looks historically constructive.