If you run an automated options strategy on NIFTY, BANK NIFTY or SENSEX, your system probably already knows how to find setups. The question it usually cannot answer is a narrower one: for this contract, right now, is there an edge that survives spread, slippage and costs?
At AION Analytics (India) we built Edge Engine to answer exactly that question and nothing else. It is a hosted API. You ask about a contract, it tells you whether it is tradable, and your own code decides what to do with that. This post shows how to wire it in as a fail-closed gate in about twenty lines of Python.
What it is, in one paragraph
Edge Engine assesses every live, unexpired NIFTY, BANK NIFTY and SENSEX contract continuously and returns a small, consistent record for each one: tradable, a signed edge_score, a conservative lower_bound, and as_of, the time of the exchange tick behind the reading. It does not place orders, never sees your positions or account, and never starts a trade. It can only stop one. The most common answer is "no", and that is the system working.
The endpoints
Base URL: https://aionanalytics.in/api/v1/edge. Your key goes in the X-API-Key header.
| Endpoint | What it returns |
|---|---|
GET /underlyings |
Which indices are live and which are coming soon. No key needed. |
GET /snapshot?underlying=nifty |
Readings for every live, unexpired contract of that index. Add &symbol= for one contract. |
POST /predict |
The same, for clients that prefer JSON: {"underlying": "banknifty", "symbol": null}
|
You can try the first one right now without a key:
curl https://aionanalytics.in/api/v1/edge/underlyings
What a real response looks like
This is from the NIFTY session on 29 September 2026 at 11:39 IST, shortened to three of the 98 contracts returned:
{
"model": "edge-engine",
"underlying": "nifty",
"count": 98,
"tradable_count": 0,
"status": "ok",
"contracts": [
{"symbol": "NFONIFTY26SEP22200CE", "instrument": "CE", "strike": 22200.0,
"expiry": "2026-09-29", "tradable": false,
"edge_score": 0.3759, "lower_bound": -4.254,
"as_of": "2026-09-29T06:09:20.787000+00:00"},
{"symbol": "NFONIFTY26O0622300PE", "instrument": "PE", "strike": 22300.0,
"expiry": "2026-10-06", "tradable": false,
"edge_score": 1.1425, "lower_bound": -0.227,
"as_of": "2026-09-29T06:09:18.402000+00:00"},
{"symbol": "NFONIFTY26SEPFUT", "instrument": "FUT", "strike": null,
"expiry": "2026-09-29", "tradable": false,
"edge_score": 3.8682, "lower_bound": -3.012,
"as_of": "2026-09-29T06:09:21.055000+00:00"}
],
"usage": {"daily_count": 12, "daily_cap": 166}
}
All three contracts have a positive edge_score, and all three are tradable: false. That is the interesting part. A positive score with a negative lower bound means the conservative estimate does not clear, so there is nothing to act on. A dashboard that showed you only the score would have shown three green numbers. On that day, no NIFTY contract was tradable at all.
The gate, in Python
The pattern we recommend is simple: your strategy proposes a contract, Edge Engine is asked once, and anything other than a clear "yes" means stand aside.
import os
from datetime import datetime, timezone, timedelta
import requests
EDGE = "https://aionanalytics.in/api/v1/edge/snapshot"
KEY = os.environ["AION_EDGE_KEY"]
MAX_AGE = timedelta(minutes=5) # your own freshness tolerance
def edge_allows(symbol: str, index: str = "nifty") -> bool:
try:
r = requests.get(EDGE, params={"underlying": index, "symbol": symbol},
headers={"X-API-Key": KEY}, timeout=5)
except requests.RequestException:
return False # unreachable: fail closed
if r.status_code != 200:
return False # 401, 403, 404, 429: fail closed
rows = r.json().get("contracts", [])
if not rows:
return False # no fresh reading: fail closed
c = rows[0]
as_of = datetime.fromisoformat(c["as_of"])
if datetime.now(timezone.utc) - as_of > MAX_AGE:
return False # older than you accept: fail closed
return bool(c["tradable"])
Then, in your order path:
if not edge_allows("NFONIFTY26O0622300PE"):
log.info("edge gate: stand aside")
return
place_order(...) # your own risk checks and order logic
Every failure path returns False. That is deliberate. A network blip, an expired key, a contract that has gone quiet, or a reading older than your tolerance all mean the same thing: you do not have an answer, so you do not trade on one.
Things worth knowing before you integrate
Gate on tradable, not on the score. The score is there so you can see why. tradable is the decision, and it already accounts for the lower bound and for how much history sits behind the reading.
Stale readings are never served. A contract only appears if its last exchange tick is under 15 minutes old. Outside market hours you get "status": "no_fresh_readings" and an empty list, not yesterday's numbers.
Poll once a minute per index. Readings look five minutes ahead and the server refreshes them continuously. Polling faster returns the same answer and uses up your daily cap.
Compare scores within one index only. SENSEX has a shorter live history than NIFTY, so it says "no edge" more often. That is honest, not broken.
Errors are boring on purpose. 401 missing key, 403 key not active or index not on your plan, 404 unknown contract or no fresh reading, 429 daily cap reached. All arrive as {"error": {"code": ..., "message": ...}}.
Pricing
- Single index: ₹10,000 a month, NIFTY, BANK NIFTY or SENSEX, 166 requests a day.
- All indices: ₹25,000 a month, all three, 500 requests a day. Opening shortly; write to contact@aionanalytics.in to reserve.
- Commodities: gold, silver and crude on MCX are coming soon.
There is no free trial. Instead, if it is not useful to you, ask for a full refund within 2 days of your first payment. Two days of live market hours is enough to see whether the gate behaves the way your strategy needs.
What it is not
It is not a strategy and not a signal service. It does not tell you what to buy or sell, it will not invent an edge to fill a quiet session, and an edge is a statement about many occurrences, not a promise about your next trade. It is informational only, not investment advice, and AION is not a SEBI-registered Research Analyst or Investment Adviser. We publish what the system does and what it returns. We do not publish how it is built.
Get started
- Check what is live:
curl https://aionanalytics.in/api/v1/edge/underlyings - Read the full field reference and error list on the Edge Engine page.
- Subscribe from the same page and drop the gate above into your order path.
If anything in the docs is unclear, there is a feedback box at the bottom of the Edge Engine page that goes straight to the team.
Top comments (0)