Version

AION India Event Intelligence

Structured event intelligence for Indian financial markets.

AION India Event Intelligence converts a single Indian financial headline into structured event, sector, stakeholder, and meta-factor output. It answers "what happened, which sectors are affected, and who gains or loses" โ€” not just "is this positive or negative."

Formerly aion-news-to-signal.

Why This Exists

Most financial NLP tools stop at sentiment polarity. Indian markets require more: NSE sectors move differently on the same RBI announcement. A crude oil shock lifts energy but pressures aviation. A weather event cascades through agriculture, logistics, and food retail. AION India Event Intelligence models these causal chains explicitly through a curated Indian-market event taxonomy, sector-level propagation rules, and meta-factor decomposition covering interest rates, crude oil, rupee pressure, risk sentiment, and liquidity. This is not sentiment. It is structured event intelligence.

Supported Production Access

AION India Event Intelligence is exposed as a managed API.

  • POST https://api.aiondashboard.site/v1/analyze
  • header:
    • X-API-Key: <key>

Example:

curl -X POST "https://api.aiondashboard.site/v1/analyze" \
  -H "Content-Type: application/json" \
  -H "X-API-Key: <key>" \
  -d '{"headline":"RBI hikes repo rate by 25 bps"}'

Python example:

import requests

resp = requests.post(
    "https://api.aiondashboard.site/v1/analyze",
    headers={"X-API-Key": "<key>"},
    json={"headline": "RBI hikes repo rate by 25 bps"},
    timeout=30,
)
resp.raise_for_status()
print(resp.json()["sector_vector"])

Output Contract

{
  "headline": "string",
  "event": "string|null",
  "confidence": "float",
  "vix_regime": "string",
  "sector_vector": {},
  "top_positive_sectors": {},
  "top_negative_sectors": {},
  "sector_directional_bias": {
    "positive_bias": [],
    "negative_bias": []
  },
  "stakeholder_views": {},
  "raw_assignment": {
    "resolved_event_id": "string|null",
    "cause_effect_rule_id": "string|null",
    "weather_triggered": "bool"
  }
}

Canonical key note:

  • sector_vector is the full scored sector impact map
  • there is no sector_impacts key in the public output contract

Quota And Access

  • Free tier: limited monthly requests via API key
  • The public Hugging Face Space is a demo environment with rate limits and cold starts โ€” it is not intended for production usage
  • Production access requires an API key from the AION dashboard
  • Requests are metered; over-quota requests are rejected

Deployment Model

AION India Event Intelligence is an API-first system. The model weights are not publicly distributed. Production access is through the managed AION API only. The PyPI package (aion-news-to-signal) provides client-side tooling and MCP server entrypoints โ€” it does not bundle model weights.

Runtime Release Metadata

  • Managed API / HF Space runtime: deterministic rule-engine updates are served through the hosted AION API.
  • Self-hosted package artifact: v1.0.0.post2-self-hosted on GitHub Releases.
  • Runtime fix scope: geopolitical conflict, geopolitical oil-supply shock, commercial LPG/fuel price policy changes, and low-confidence out-of-domain fallback suppression.
  • Model weights: unchanged. No DistilBERT retraining was performed for this release.

Licensing Note

The public client/code surface for AION India Event Intelligence is distributed under AGPL-3.0-or-later. Some deterministic taxonomy rule files in the main GitHub repository are governed separately under Business Source License 1.1 with a future AGPL change date. See the repository LICENSE and LICENSE.BSL11 files for the exact split.

Use This For

  • Indian financial headline reasoning
  • event classification with sector propagation
  • stakeholder-level interpretation
  • dashboard, agent, and research integrations

Do Not Use This For

  • blind execution
  • broker integration by itself
  • compliance or suitability decisions
  • unsupported local inference workflows

Next Steps

  1. Try the demo: https://huggingface.co/spaces/AION-Analytics/aion-news-to-signal
  2. Get an API key: https://dashboard.aiondashboard.site/access/register
  3. View full documentation: https://dashboard.aiondashboard.site/models/news-to-signal
  4. Integrate via MCP: https://dashboard.aiondashboard.site/models/news-to-signal

Cross Links

  • API gateway:
    • https://dashboard.aiondashboard.site/systems/api-gateway
  • GitHub repository:
    • https://github.com/AION-Analytics/aion-news-to-signal
  • PyPI package:
    • https://pypi.org/project/aion-news-to-signal/
  • website model page:
    • https://dashboard.aiondashboard.site/models/news-to-signal
  • registration:
    • https://dashboard.aiondashboard.site/access/register
  • live demo space:
    • https://huggingface.co/spaces/AION-Analytics/aion-news-to-signal

Citation

@software{aion_india_event_intelligence_2026,
  author = {AION Analytics},
  title = {AION India Event Intelligence},
  year = {2026},
  url = {https://huggingface.co/AION-Analytics/aion-news-to-signal},
}
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support