---
title: "A Firehose of the Web, Already Enriched with Entities, Tickers, Sentiments, Ontologies, and Locations"
description: "NOSIBLE Firehose is a live websocket feed of crawled web documents, with entities, tickers, locations, ontologies, sentiment and embeddings on every snippet."
url: "https://nosible.com/blog/a-firehose-of-the-web-already-enriched"
---

[Information Retrieval](https://nosible.com/blog/tag/information-retrieval)

[Web Search](https://nosible.com/blog/tag/web-search)

# A Firehose of the Web, Already Enriched with Entities, Tickers, Sentiments, Ontologies, and Locations

[Stuart Reid](https://www.linkedin.com/in/stuartgordonreid/)

2026-11-17

21 min read

Copy as Markdown

![Documents per hour for Nvidia and Jensen Huang in green above sentiment in red and the NVDA share price in white, from 7 to 13 November 2026, with the volume spike and the price drop lining up.](https://nosible.com/images/2026/11/firehose-nvidia-hero-3640x1560.png)

Today we're sharing that Firehose, our lowest latency product, is officially in preview. Firehose is a core part of our plan to build ***complementary*** products along the latency spectrum with different levels of intelligence.

Firehose is a live feed of the web documents we crawl and enrich. We split every document into snippets, the passages of its text, and enrich every snippet on its own, so the enrichment points at the exact passage and not only at the page.

Each snippet carries:

- **Named entities:** the people, organisations, places and products it names, each with a type and confidence, so you can see who or what it is about.
- **Locations:** the places the passage is about, geocoded to a continent, region or country, so you can filter by where something happened.
- **Tickers:** the listed companies it matches, with identifiers such as ISIN, LEI and FIGI, plus sector and exchange, so you can route it to a position.
- **Ontologies:** the topics it covers, scored by probability across 15 [ontology families](https://docs.nosible.com/endpoints/world-ontology-catalog), such as GICS sectors, emotions and event classes.
- **Sentiment:** real-time classification of the passage as negative, neutral or positive using [model v1.2](https://huggingface.co/NOSIBLE/financial-sentiment-v1.2-base), with the probability of each.
- **Forward-looking:** real-time classification of whether it describes something yet to happen (forward) or not (not-forward), by [model v1.2](https://huggingface.co/NOSIBLE/forward-looking-v1.2-base), with the probability of each.
- **Embedding:** the passage's own 768-dimension vector, so you can run semantic search or clustering on a passage instead of a whole page.

The page title and description are enriched the same way, and the document carries the metadata, the full text and a roll-up of everything its snippets say.

Documents are delivered over a real-time websocket, and consumers can submit filters that run server side. So you can listen to everything or just the documents that match certain metadata, topics, tickers, geolocations, named entities and ontologies. For example, stream just the documents tagged to Nvidia or Jensen Huang.

Let's do that. Our story begins 8 November, when Nvidia's chief executive, Jensen Huang, admitted to an affair. First we'll connect to the websocket and submit a server side filter. Then we'll see how documents tagged with Nvidia or Jensen Huang changed this week, and how their sentiment compares with the share price.

Table of contents [Why You Actually Want a Web-Scale Firehose](https://nosible.com/blog/a-firehose-of-the-web-already-enriched#why-you-actually-want-a-web-scale-firehose) [Where Does NOSIBLE Firehose Sit in the Stack?](https://nosible.com/blog/a-firehose-of-the-web-already-enriched#where-does-nosible-firehose-sit-in-the-stack) [How the Firehose Works: Connect and Receive](https://nosible.com/blog/a-firehose-of-the-web-already-enriched#how-the-firehose-works-connect-and-receive) [Filters and Options for Your Subscription](https://nosible.com/blog/a-firehose-of-the-web-already-enriched#filters-and-options-for-your-subscription) [Replaying Nvidia and Jensen Huang Through the Firehose](https://nosible.com/blog/a-firehose-of-the-web-already-enriched#replaying-nvidia-and-jensen-huang-through-the-firehose) [Firehose is Best for Mid-Frequency Strategies](https://nosible.com/blog/a-firehose-of-the-web-already-enriched#firehose-is-best-for-mid-frequency-strategies) [Get Early Access to the Firehose in Preview](https://nosible.com/blog/a-firehose-of-the-web-already-enriched#get-early-access-to-the-firehose-in-preview)

## Why You Actually Want a Web-Scale Firehose

Your news feed publishes a story after it happens, and that is when you find out. If it breaks in a local foreign-language publication, the gap can be anywhere from a day to a week, and anyone who saw it elsewhere can already act on it.

The truth is that you don't actually want a *news* feed. What you actually want is a feed of the worldwide web. Here's the simple reason why:

Say 10,000 websites cover a story. One publishes first and the other 9,999 publish after it. The chance that your *news* feed is the first is about 1 in 10,000, so almost every time the story is already out before your feed has picked it up.

Firehose watches hundreds of thousands of websites, including local publications in 95 languages. The first one to publish is far more likely to be one of them, so you learn about the story much sooner, wherever in the world it appears.

## Where Does NOSIBLE Firehose Sit in the Stack?

The NOSIBLE stack has four levels: Firehose, Search, World and Atlas. Firehose is the fastest. Each level after it is slower and more intelligent than the one before.

Traders know the same trade-off from market data. Packets carry raw messages, ticks record single trades and quote changes as they happen, intraday bars summarise a period such as an hour, and end-of-day data summarises a whole session. Each of these is slower than the one before and tells you more.

Funds trade at different frequencies, so they use different levels of market data. A daily allocator gains nothing from ticks. A fund that reacts within seconds cannot use end-of-day bars, because they say nothing about the last second. The most intelligent level cannot also be the fastest, so each suits a different decision.

![A staircase chart of the NOSIBLE stack. Speed runs left to right and intelligence runs bottom to top. Firehose, a near-real-time data feed of what we crawl, is at the bottom left. Search, a search engine for dated web documents, sits above it, then World, a point-in-time database of worldwide events. Atlas, a point-in-time knowledge graph built from World, is at the top right. Documents arrive as soon as they are crawled, results as soon as they are indexed, events as soon as they are detected and facts as soon as they are corroborated.](https://nosible.com/images/2026/11/firehose-stack-spectrum.svg)

Higher up the chart, more is folded into each item, so it is more intelligent. The cost is time: an item is ready after its inputs arrive.

Search, World and Atlas all answer when you ask. Firehose is the one level that pushes: a document is sent to your client when it is ready, so nothing has to poll on a timer and wait up to the length of the timer for an answer.

Firehose also does all the work before delivery that needs only one document, and it does it on every snippet: embeddings, ontologies, locations, real-time sentiment and real-time forward-looking classification and, for English documents, entities and tickers. Every subscriber receives the same result.

## How the Firehose Works: Connect and Receive

Firehose is one WebSocket. You send two things, your API key and one subscribe request. The server sends back two confirmations and then a stream of documents. Your application needs WebSocket support, JSON parsing, Zstandard decompression, and durable storage for its resume cursor. This is a quick example; for everything else, see our [API documentation](https://docs.nosible.com/endpoints/firehose-connect) or contact the team.

### 1. Connect with your API key

Open `wss://firehose.nosible.world/v2.0/firehose` with your API key in an `X-Api-Key` or `Authorization: Bearer` header, then wait for `authenticated`. If it cannot set headers, send `{"type":"auth","api_key":"YOUR_API_KEY"}` as a text message within 3 seconds. Your key must be approved for Firehose access.

// authenticated (server sends)

Expand all Collapse all Copy

{

type, can_replay, time_ms

}

"type": "authenticated",

"can_replay": true,

"time_ms": 1794303700000

}

### 2. Send a subscribe request

Within 10 seconds of `authenticated`, send one subscribe request. It asks for every document that names Nvidia or Jensen Huang, its chief executive from our story. Different entities in one filter combine with OR, so a single subscription follows both:

// subscribe request (you send)

Expand all Collapse all Copy

{

type, subscription, resume, filters

}

"type": "subscribe",

"subscription": "nvidia-and-jensen-huang",

"resume": "latest",

"filters"

:

{

entities

}

"entities"

:

[

2 items

]

{

label, name

},

"label": "ORG",

"name": "Nvidia"

},

{

label, name

}

"label": "PERSON",

"name": "Jensen Huang"

}

]

}

}

`subscription` names your feed, `resume` says where to start, `filters` pick the documents, and `fields` (omitted) picks what you receive. The server replies with `subscribed`, echoing your name and start ( `source` ) and a `cursor`, then sends documents. A name takes 1 to 64 letters, digits, hyphens or underscores, and reusing an active name with the same key replaces the previous connection.

### 3. Receive documents

After `subscribed`, each matching document arrives as a binary message, and a text `heartbeat` carrying your cursor arrives every 15 seconds, even when nothing matches. Every message the server sends includes `time_ms`, the time it was prepared in UTC Unix milliseconds. A document message is a plain JSON header you can read without decompressing, then the zstd-compressed document: the bytes `NFH1`, a 4-byte big-endian header length, the header, and one zstd frame. Our [Firehose reference](https://docs.nosible.com/endpoints/firehose-document) has Python and Node decoders.

Below is a hypothetical document for this subscription, inspired by NOSIBLE World's 10 November 2026 Nvidia story. The article and publisher are invented, the tickers and topics come from the event, and every snippet's embeddings, entities, locations, sentiment, forward-looking scores and ontologies are synthetic. The header:

// header (plain JSON)

Expand all Collapse all Copy

{

cursor, arrived_ms, time_ms, facets

}

"cursor": "eyJ2IjoxLC4uLn0",

"arrived_ms": 1794303761000,

"time_ms": 1794303761100,

"facets"

:

{

10 keys

}

"hash": "9IvHQRPbpKOynMQLU4Z5ugBn",

"netloc": "news.example.com",

"published_timestamp": "2026-11-10T09:15:00+00:00",

"visited_timestamp": "2026-11-10T09:42:07+00:00",

"languages"

:

{

en

},

"en": 0.999

},

"entities"

:

{

GPE, ORG, PERSON

},

"GPE"

:

[

1 item

],

{

6 keys

}

"name": "US",

"count": 1,

"mention_count": 1,

"min_conf": 0.9941,

"max_conf": 0.9941,

"mean_conf": 0.9941

}

],

"ORG"

:

[

1 item

],

{

6 keys

}

"name": "Nvidia",

"count": 4,

"mention_count": 4,

"min_conf": 0.999,

"max_conf": 0.9998,

"mean_conf": 0.9994

}

],

"PERSON"

:

[

2 items

]

{

6 keys

},

"name": "Jensen Huang",

"count": 2,

"mention_count": 2,

"min_conf": 0.9999,

"max_conf": 1,

"mean_conf": 1

},

{

6 keys

}

"name": "Huang",

"count": 1,

"mention_count": 1,

"min_conf": 0.9982,

"max_conf": 0.9982,

"mean_conf": 0.9982

}

]

},

"locations"

:

[

1 item

],

{

8 keys

}

"lat": 39.76,

"lng": -98.5,

"continent": "North America",

"region": "North America",

"country": "United States",

"source": "ner",

"matched_on": "US",

"entity_type": "GPE"

}

],

"ontologies"

:

[

7 items

],

{

8 keys

},

"ontology": "nosible",

"category": "Reputational Risk",

"subcategory": "Executive Misconduct",

"event": "Alleged Sexual Harassment",

"count": 3,

"min_prob": 0.741,

"max_prob": 0.9602,

"mean_prob": 0.8613

},

{

6 keys

},

"ontology": "media_frame",

"label": "Responsibility",

"count": 2,

"min_prob": 0.8933,

"max_prob": 0.9115,

"mean_prob": 0.9024

},

{

6 keys

},

"ontology": "ekman7_emotion",

"label": "surprise",

"count": 1,

"min_prob": 0.712,

"max_prob": 0.712,

"mean_prob": 0.712

},

{

8 keys

},

"ontology": "nosible",

"category": "Technical Analysis",

"subcategory": "Price Action",

"event": "Share Price Decreased",

"count": 1,

"min_prob": 0.9611,

"max_prob": 0.9611,

"mean_prob": 0.9611

},

{

9 keys

},

"ontology": "gics",

"industry": "Semiconductors & Semiconductor Equipment",

"industry_group": "Semiconductors & Semiconductor Equipment",

"sector": "Information Technology",

"sub_industry": "Semiconductors",

"count": 1,

"min_prob": 0.9,

"max_prob": 0.9,

"mean_prob": 0.9

},

{

8 keys

},

"ontology": "nosible",

"category": "Corporate Governance",

"subcategory": "Executive Leadership",

"event": "Chief Executive Officer Departed",

"count": 1,

"min_prob": 0.5812,

"max_prob": 0.5812,

"mean_prob": 0.5812

},

{

6 keys

}

"ontology": "ekman7_emotion",

"label": "fear",

"count": 1,

"min_prob": 0.6204,

"max_prob": 0.6204,

"mean_prob": 0.6204

}

],

"tickers"

:

[

1 item

],

{

15 keys

}

"ticker": "NVDA.US",

"companyv5_id": "c5-1b5479ea345b",

"name": "NVIDIA Corporation",

"website": "https://nvidia.com",

"wiki_qid": "Q182477",

"isin": "US67066G1040",

"lei": "549300S4KLFTLO7GSQ80",

"figi": "BBG000BBJQV0",

"sector": "Information Technology",

"industry": "Semiconductors & Semiconductor Equipment",

"country_iso": "US",

"exchange": "NASDAQ",

"mic": "XNAS",

"score": 0.995505,

"why"

:

[

1 item

]

{

entity, score, chain

}

"entity": "Jen-Hsun \"Jensen\" Huang",

"score": 0.081243,

"chain"

:

[

1 item

]

{

subject, predicate, object, known_as_of

}

"subject": "Jen-Hsun \"Jensen\" Huang",

"predicate": "ceo_of",

"object": "Nvidia",

"known_as_of": "2010-01-08"

}

]

}

]

}

],

"has_tickers": true

}

}

The header says where the document sits in the stream. Its `facets` are flat copies of selected document fields, with the same names, so you can filter further, for example on tickers, without decompressing. `cursor` is the position to save. `arrived_ms` is when it reached our stream, not when published. `facets.hash`, `facets.published_timestamp` and `facets.visited_timestamp` copy `request.hash`, `page.published_timestamp` and `page.visited_timestamp`. The hash and the visited timestamp identify the document, so you can drop duplicates.

This is the document inside the zstd part, from the same sample:

// document (zstd frame)

Expand all Collapse all Copy

{

13 keys

}

"request"

:

{

11 keys

},

"raw_url": "https://news.example.com/tech/huang-admits-affair",

"url": "https://news.example.com/tech/huang-admits-affair",

"hash": "9IvHQRPbpKOynMQLU4Z5ugBn",

"geo": null,

"proxy": null,

"scheme": "https",

"netloc": "news.example.com",

"prefix": "news",

"domain": "example",

"suffix": "com",

"path": "/tech/huang-admits-affair"

},

"page"

:

{

16 keys

},

"title": "Nvidia chief Jensen Huang admits affair",

"description": "Nvidia's chief executive has admitted to an affair.",

"author": "News Desk",

"published": "2026-11-10",

"modified": "",

"visited": "2026-11-10",

"certain": true,

"published_timestamp": "2026-11-10T09:15:00+00:00",

"modified_timestamp": "",

"visited_timestamp": "2026-11-10T09:42:07+00:00",

"entities"

:

{

ORG, PERSON

},

"ORG"

:

[

1 item

],

{

name, mention_count, confidence

}

"name": "Nvidia",

"mention_count": 1,

"confidence": 0.999

}

],

"PERSON"

:

[

1 item

]

{

name, mention_count, confidence

}

"name": "Jensen Huang",

"mention_count": 1,

"confidence": 0.9999

}

]

},

"locations": [ ],

"tickers": [ ],

"signals"

:

{

2 keys

},

"financial_sentiment"

:

{

label, probabilities

},

"label": "negative",

"probabilities"

:

{

negative, neutral, positive

}

"negative": 0.9803,

"neutral": 0.0187,

"positive": 0.001

}

},

"forward_looking"

:

{

label, probabilities

}

"label": "not-forward",

"probabilities"

:

{

forward, not-forward

}

"forward": 0.1286,

"not-forward": 0.8714

}

}

},

"ontologies"

:

[

2 items

],

{

ontology, category, subcategory, event, probability

},

"ontology": "nosible",

"category": "Reputational Risk",

"subcategory": "Executive Misconduct",

"event": "Alleged Sexual Harassment",

"probability": 0.9602

},

{

ontology, label, probability

}

"ontology": "media_frame",

"label": "Responsibility",

"probability": 0.9115

}

],

"embeddings"

:

{

embeddinggemma-2

}

"embeddinggemma-2": "H/7F/5YAogHoAewCBwGk/R/+…" ( 2,048 characters)

}

},

"languages"

:

{

en

},

"en": 0.999

},

"full_text"

:

"Example News - Technology Nvidia chief Jensen Hu…"

( 473 characters)

,

"Example News - Technology\nNvidia chief Jensen Huang admits affair\nBy News Desk\nNvidia chief Jensen Huang told staff on 8 November that he had been unfaithful to his wife.\nNvidia shares fell more than 4 percent as US officials weighed national security questions.\nNvidia's board said it had opened an internal investigation and asked an outside law firm to review the matter.\nAnalysts said investors would look to Huang for a statement on his future before the market opens.",

"snippets"

:

{

4 keys

},

"FvMVztP4NAuli6lYU4Pk67Ey"

:

{

15 keys

},

"url_hash": "9IvHQRPbpKOynMQLU4Z5ugBn",

"snippet_hash": "FvMVztP4NAuli6lYU4Pk67Ey",

"prev_snippet_hash": null,

"next_snippet_hash": "k4wlVv963sQInODe8ZcVnVEK",

"content": "Nvidia chief Jensen Huang told staff on 8 November that he had been unfaithful to his wife.",

"words": "nvidia chief jensen huang told staff 8 novemb unfaith wife",

"language": "en",

"statistics"

:

{

sentences, words, characters, links

},

"sentences": 1,

"words": 17,

"characters": 91,

"links": 1

},

"links"

:

{

Jensen Huang

},

"Jensen Huang": "https://news.example.com/people/jensen-huang"

},

"entities"

:

{

ORG, PERSON

},

"ORG"

:

[

1 item

],

{

name, mention_count, confidence

}

"name": "Nvidia",

"mention_count": 1,

"confidence": 0.9995

}

],

"PERSON"

:

[

1 item

]

{

name, mention_count, confidence

}

"name": "Jensen Huang",

"mention_count": 1,

"confidence": 1

}

]

},

"locations": [ ],

"tickers": [ ],

"signals"

:

{

2 keys

},

"financial_sentiment"

:

{

label, probabilities

},

"label": "negative",

"probabilities"

:

{

negative, neutral, positive

}

"negative": 0.9712,

"neutral": 0.0273,

"positive": 0.0015

}

},

"forward_looking"

:

{

label, probabilities

}

"label": "not-forward",

"probabilities"

:

{

forward, not-forward

}

"forward": 0.0598,

"not-forward": 0.9402

}

}

},

"ontologies"

:

[

2 items

],

{

ontology, category, subcategory, event, probability

},

"ontology": "nosible",

"category": "Reputational Risk",

"subcategory": "Executive Misconduct",

"event": "Alleged Sexual Harassment",

"probability": 0.8826

},

{

ontology, label, probability

}

"ontology": "ekman7_emotion",

"label": "surprise",

"probability": 0.712

}

],

"embeddings"

:

{

embeddinggemma-2

}

"embeddinggemma-2": "GAOgACoAnf31AKr/fQHjAfv9…" ( 2,048 characters)

}

},

"k4wlVv963sQInODe8ZcVnVEK"

:

{

15 keys

},

"url_hash": "9IvHQRPbpKOynMQLU4Z5ugBn",

"snippet_hash": "k4wlVv963sQInODe8ZcVnVEK",

"prev_snippet_hash": "FvMVztP4NAuli6lYU4Pk67Ey",

"next_snippet_hash": "sXJsbSUufAhJvvVsWLNJPJ3p",

"content": "Nvidia shares fell more than 4 percent as US officials weighed national security questions.",

"words": "nvidia share fell more 4 percent us offici weigh nation secur question",

"language": "en",

"statistics"

:

{

sentences, words, characters, images

},

"sentences": 1,

"words": 14,

"characters": 91,

"images": 1

},

"images"

:

[

1 item

],

"https://news.example.com/images/nvda-share-price.png"

],

"entities"

:

{

GPE, ORG

},

"GPE"

:

[

1 item

],

{

name, mention_count, confidence

}

"name": "US",

"mention_count": 1,

"confidence": 0.9941

}

],

"ORG"

:

[

1 item

]

{

name, mention_count, confidence

}

"name": "Nvidia",

"mention_count": 1,

"confidence": 0.9998

}

]

},

"locations"

:

[

1 item

],

{

8 keys

}

"lat": 39.76,

"lng": -98.5,

"continent": "North America",

"region": "North America",

"country": "United States",

"source": "ner",

"matched_on": "US",

"entity_type": "GPE"

}

],

"tickers": [ ],

"signals"

:

{

2 keys

},

"financial_sentiment"

:

{

label, probabilities

},

"label": "negative",

"probabilities"

:

{

negative, neutral, positive

}

"negative": 0.954,

"neutral": 0.0431,

"positive": 0.0029

}

},

"forward_looking"

:

{

label, probabilities

}

"label": "not-forward",

"probabilities"

:

{

forward, not-forward

}

"forward": 0.1197,

"not-forward": 0.8803

}

}

},

"ontologies"

:

[

2 items

],

{

ontology, category, subcategory, event, probability

},

"ontology": "nosible",

"category": "Technical Analysis",

"subcategory": "Price Action",

"event": "Share Price Decreased",

"probability": 0.9611

},

{

6 keys

}

"ontology": "gics",

"industry": "Semiconductors & Semiconductor Equipment",

"industry_group": "Semiconductors & Semiconductor Equipment",

"sector": "Information Technology",

"sub_industry": "Semiconductors",

"probability": 0.9

}

],

"embeddings"

:

{

embeddinggemma-2

}

"embeddinggemma-2": "nf9nAJX+kgBh/zkCaQBF/z39…" ( 2,048 characters)

}

},

"sXJsbSUufAhJvvVsWLNJPJ3p"

:

{

15 keys

},

"url_hash": "9IvHQRPbpKOynMQLU4Z5ugBn",

"snippet_hash": "sXJsbSUufAhJvvVsWLNJPJ3p",

"prev_snippet_hash": "k4wlVv963sQInODe8ZcVnVEK",

"next_snippet_hash": "R0HiUrvCJ5gW8XzNqoVnyTrk",

"content": "Nvidia's board said it had opened an internal investigation and asked an outside law firm to review the matter.",

"words": "nvidia s board said open intern investig ask outsid law firm review matter",

"language": "en",

"statistics"

:

{

sentences, words, characters, lists

},

"sentences": 1,

"words": 19,

"characters": 111,

"lists": 1

},

"lists"

:

[

1 item

],

[

3 items

]

"Outside law firm appointed",

"Audit committee briefed",

"Review of conduct policies opened"

]

],

"entities"

:

{

ORG

},

"ORG"

:

[

1 item

]

{

name, mention_count, confidence

}

"name": "Nvidia",

"mention_count": 1,

"confidence": 0.9993

}

]

},

"locations": [ ],

"tickers": [ ],

"signals"

:

{

2 keys

},

"financial_sentiment"

:

{

label, probabilities

},

"label": "negative",

"probabilities"

:

{

negative, neutral, positive

}

"negative": 0.8211,

"neutral": 0.1672,

"positive": 0.0117

}

},

"forward_looking"

:

{

label, probabilities

}

"label": "not-forward",

"probabilities"

:

{

forward, not-forward

}

"forward": 0.279,

"not-forward": 0.721

}

}

},

"ontologies"

:

[

2 items

],

{

ontology, category, subcategory, event, probability

},

"ontology": "nosible",

"category": "Reputational Risk",

"subcategory": "Executive Misconduct",

"event": "Alleged Sexual Harassment",

"probability": 0.741

},

{

ontology, label, probability

}

"ontology": "media_frame",

"label": "Responsibility",

"probability": 0.8933

}

],

"embeddings"

:

{

embeddinggemma-2

}

"embeddinggemma-2": "8v+wAML8JwDd/kD/7/3G/9MA…" ( 2,048 characters)

}

},

"R0HiUrvCJ5gW8XzNqoVnyTrk"

:

{

14 keys

}

"url_hash": "9IvHQRPbpKOynMQLU4Z5ugBn",

"snippet_hash": "R0HiUrvCJ5gW8XzNqoVnyTrk",

"prev_snippet_hash": "sXJsbSUufAhJvvVsWLNJPJ3p",

"next_snippet_hash": null,

"content": "Analysts said investors would look to Huang for a statement on his future before the market opens.",

"words": "analyst said investor would look huang statement futur befor market open",

"language": "en",

"statistics"

:

{

sentences, words, characters

},

"sentences": 1,

"words": 17,

"characters": 98

},

"entities"

:

{

PERSON

},

"PERSON"

:

[

1 item

]

{

name, mention_count, confidence

}

"name": "Huang",

"mention_count": 1,

"confidence": 0.9982

}

]

},

"locations": [ ],

"tickers": [ ],

"signals"

:

{

2 keys

},

"financial_sentiment"

:

{

label, probabilities

},

"label": "neutral",

"probabilities"

:

{

negative, neutral, positive

}

"negative": 0.3518,

"neutral": 0.632,

"positive": 0.0162

}

},

"forward_looking"

:

{

label, probabilities

}

"label": "forward",

"probabilities"

:

{

forward, not-forward

}

"forward": 0.912,

"not-forward": 0.088

}

}

},

"ontologies"

:

[

2 items

],

{

ontology, category, subcategory, event, probability

},

"ontology": "nosible",

"category": "Corporate Governance",

"subcategory": "Executive Leadership",

"event": "Chief Executive Officer Departed",

"probability": 0.5812

},

{

ontology, label, probability

}

"ontology": "ekman7_emotion",

"label": "fear",

"probability": 0.6204

}

],

"embeddings"

:

{

embeddinggemma-2

}

"embeddinggemma-2": "HADB/7b+F/93/2X/wQE8/gT9…" ( 2,048 characters)

}

}

},

"statistics"

:

{

12 keys

},

"snippets": 4,

"sentences": 4,

"words": 67,

"characters": 391,

"images": 1,

"videos": 0,

"audio": 0,

"tables": 0,

"lists": 1,

"blocks": 0,

"links": 1,

"files": 0

},

"entities"

:

{

GPE, ORG, PERSON

},

"GPE"

:

[

1 item

],

{

6 keys

}

"name": "US",

"count": 1,

"mention_count": 1,

"min_conf": 0.9941,

"max_conf": 0.9941,

"mean_conf": 0.9941

}

],

"ORG"

:

[

1 item

],

{

6 keys

}

"name": "Nvidia",

"count": 4,

"mention_count": 4,

"min_conf": 0.999,

"max_conf": 0.9998,

"mean_conf": 0.9994

}

],

"PERSON"

:

[

2 items

]

{

6 keys

},

"name": "Jensen Huang",

"count": 2,

"mention_count": 2,

"min_conf": 0.9999,

"max_conf": 1,

"mean_conf": 1

},

{

6 keys

}

"name": "Huang",

"count": 1,

"mention_count": 1,

"min_conf": 0.9982,

"max_conf": 0.9982,

"mean_conf": 0.9982

}

]

},

"tickers"

:

[

1 item

],

{

15 keys

}

"ticker": "NVDA.US",

"companyv5_id": "c5-1b5479ea345b",

"name": "NVIDIA Corporation",

"website": "https://nvidia.com",

"wiki_qid": "Q182477",

"isin": "US67066G1040",

"lei": "549300S4KLFTLO7GSQ80",

"figi": "BBG000BBJQV0",

"sector": "Information Technology",

"industry": "Semiconductors & Semiconductor Equipment",

"country_iso": "US",

"exchange": "NASDAQ",

"mic": "XNAS",

"score": 0.995505,

"why"

:

[

1 item

]

{

entity, score, chain

}

"entity": "Jen-Hsun \"Jensen\" Huang",

"score": 0.081243,

"chain"

:

[

1 item

]

{

subject, predicate, object, known_as_of

}

"subject": "Jen-Hsun \"Jensen\" Huang",

"predicate": "ceo_of",

"object": "Nvidia",

"known_as_of": "2010-01-08"

}

]

}

]

}

],

"ontologies"

:

[

7 items

],

{

8 keys

},

"ontology": "nosible",

"category": "Reputational Risk",

"subcategory": "Executive Misconduct",

"event": "Alleged Sexual Harassment",

"count": 3,

"min_prob": 0.741,

"max_prob": 0.9602,

"mean_prob": 0.8613

},

{

6 keys

},

"ontology": "media_frame",

"label": "Responsibility",

"count": 2,

"min_prob": 0.8933,

"max_prob": 0.9115,

"mean_prob": 0.9024

},

{

6 keys

},

"ontology": "ekman7_emotion",

"label": "surprise",

"count": 1,

"min_prob": 0.712,

"max_prob": 0.712,

"mean_prob": 0.712

},

{

8 keys

},

"ontology": "nosible",

"category": "Technical Analysis",

"subcategory": "Price Action",

"event": "Share Price Decreased",

"count": 1,

"min_prob": 0.9611,

"max_prob": 0.9611,

"mean_prob": 0.9611

},

{

9 keys

},

"ontology": "gics",

"industry": "Semiconductors & Semiconductor Equipment",

"industry_group": "Semiconductors & Semiconductor Equipment",

"sector": "Information Technology",

"sub_industry": "Semiconductors",

"count": 1,

"min_prob": 0.9,

"max_prob": 0.9,

"mean_prob": 0.9

},

{

8 keys

},

"ontology": "nosible",

"category": "Corporate Governance",

"subcategory": "Executive Leadership",

"event": "Chief Executive Officer Departed",

"count": 1,

"min_prob": 0.5812,

"max_prob": 0.5812,

"mean_prob": 0.5812

},

{

6 keys

}

"ontology": "ekman7_emotion",

"label": "fear",

"count": 1,

"min_prob": 0.6204,

"max_prob": 0.6204,

"mean_prob": 0.6204

}

],

"locations"

:

[

1 item

],

{

8 keys

}

"lat": 39.76,

"lng": -98.5,

"continent": "North America",

"region": "North America",

"country": "United States",

"source": "ner",

"matched_on": "US",

"entity_type": "GPE"

}

],

"has_tickers": true,

"signals"

:

{

2 keys

},

"financial_sentiment"

:

{

negative, neutral, positive

},

"negative"

:

{

count, min, max, mean

},

"count": 4,

"min": 0.3518,

"max": 0.9803,

"mean": 0.8157

},

"neutral"

:

{

count, min, max, mean

},

"count": 1,

"min": 0.0187,

"max": 0.632,

"mean": 0.1777

},

"positive"

:

{

count, min, max, mean

}

"count": 0,

"min": 0.001,

"max": 0.0162,

"mean": 0.0067

}

},

"forward_looking"

:

{

forward, not-forward

}

"forward"

:

{

count, min, max, mean

},

"count": 1,

"min": 0.0598,

"max": 0.912,

"mean": 0.2998

},

"not-forward"

:

{

count, min, max, mean

}

"count": 4,

"min": 0.088,

"max": 0.9402,

"mean": 0.7002

}

}

},

"complete": true

}

A full document has 13 top-level keys:

- `request` holds the page address and its hash.
- `page` holds the title, description, publish and crawl times and a 768-dimension embedding. The page is enriched like everything else: `page.entities`, `page.locations`, `page.tickers`, `page.ontologies` and `page.signals` describe the title and description.
- `full_text` is the page text, and `snippets` splits it into passages.
- `statistics` counts what the crawler found.
- `languages`, `has_tickers` and `complete` describe the whole document. `complete` is `true` when every part has been enriched.

Every snippet is a passage of the page text with its own identity: `content`, the `words` we index, its `language`, `statistics` for sentences, words and characters, a `snippet_hash`, and the hashes of the snippets before and after it ( `prev_snippet_hash` and `next_snippet_hash` ), so you can walk the page in order. Each is enriched on its own, with the same fields as the page:

- `embeddings`: its own 768-dimension vector. Decode it from base64 into 768 little-endian signed 16-bit integers, then divide each by 10,000.
- `entities`: the people, organisations, places and products it names, each with a confidence.
- `locations`: the places it is about, geocoded.
- `tickers`: the listed companies it matches.
- `ontologies`: its topics, scored across the 15 ontology families.
- `signals`: its own sentiment (negative, neutral or positive) and forward-looking (forward or not-forward) classification, each with a label and probabilities.

So you can ask which passage is negative, which one is forward-looking, and which one names a company, a place and a topic together.

A snippet also carries the structure around its text, attached to the passage it belongs to. Each key below is present only when the snippet has one, and `statistics` counts it:

- `links`: link text mapped to its URL.
- `images`, `videos`, `audio` and `files`: URLs of the media and documents (such as PDF, Word, Excel and CSV files) beside the passage.
- `tables`: tables as column-keyed objects.
- `lists`: bulleted and numbered lists, as arrays of items.
- `blocks`: code blocks, as arrays of lines.

In the example, the first snippet has a link, the second an image and the third a list. The document-level `statistics` totals each of them.

The document-level `entities`, `ontologies`, `signals`, `tickers` and `locations` roll these up across the page and all of its snippets, so they report how often something occurred and how confident the models were:

- An entity has a `count` (the passages that mention it), a `mention_count` and the minimum, maximum and mean confidence.
- An ontology has a `count` and the minimum, maximum and mean probability.
- `signals` gives, for each sentiment and forward-looking class, how many passages carry that label (`count`) and the minimum, maximum and mean probability of that class. In the example, four of the five passages, counting the page, are negative and the mean probability of negative is 0.82.

Accept documents with or without `signals`. If your request sets `fields`, you receive only the parts you asked for, plus `request` and `page.visited_timestamp`, and the facets still describe the full document.

## Filters and Options for Your Subscription

You set

It keeps documents

For example

Topic

ontologies

About a topic, from 15 families, above a minimum probability

For example

GICS sector "information technology" at 0.8

Place

locations

About a continent, region or country

For example

France

Entity

entities

That name a person, organisation, place or product, matched exactly

For example

Person "Jensen Huang"

Ticker

tickers

That match any ticker symbol you list

For example

AAPL, SHEL.L

Has tickers

has_tickers

That have a ticker match, or with false, none

For example

true

Language

languages

Whose highest-scoring language is one you list

For example

en, fr

Website

netloc

From a site, including its subdomains

For example

reuters.com

Fields

fields

Sends only the parts of each document you ask for

For example

page.title, tickers

Start point

resume

Sets where in the stream you begin

For example

latest, a saved cursor, or a time in the last 24 hours

Different filters combine with AND, options within one filter combine with OR, and a filter you leave out matches everything. Text matching ignores case, but entity names and ticker symbols must match in full, and properties in one clause must match the same row. Each filter takes up to 100 options, and each key can hold four subscriptions. The hash and the published and visited timestamps cannot be filtered. Setting `has_tickers` to `false` with a non-empty `tickers` filter can never match. For example, this subscription keeps English documents that matched a ticker:

// subscribe filter (you send)

Expand all Collapse all Copy

{

type, subscription, resume, filters

}

"type": "subscribe",

"subscription": "english-companies",

"resume": "latest",

"filters"

:

{

languages, has_tickers

}

"languages"

:

[

1 item

],

"en"

],

"has_tickers": true

}

}

The [subscribe reference](https://docs.nosible.com/endpoints/firehose-subscribe) has the full syntax.

NOSIBLE does not store your position for you. Save a cursor only after you have safely processed every document before it, and send it back as `resume` when you reconnect, with the same filters and fields. You can also start from a time in the last 24 hours with `{"from_time_ms": 1794940000000}`, a UTC Unix time in milliseconds, and once a replay catches up the server sends `replay_complete` and then live documents.

// resume (you send)

Expand all Collapse all Copy

{

type, subscription, resume

}

"type": "subscribe",

"subscription": "my-feed",

"resume"

:

{

cursor

}

"cursor": "LAST_SAFELY_PROCESSED_CURSOR"

}

}

Delivery is at-least-once, so always deduplicate on `request.hash` and `page.visited_timestamp`. Keep reading, because a blocked send times out after 30 seconds and can end in `slow_consumer`. Retry temporary failures such as `upstream_unavailable` with backoff and keep your cursor rather than switching to `latest`. Headers are capped at 513 KiB and decompressed documents at 16 MiB. The [errors reference](https://docs.nosible.com/endpoints/firehose-errors) lists every error code.

## Replaying Nvidia and Jensen Huang Through the Firehose

You should know

Firehose went live on 7 November and its schema has evolved since, so older documents in a replay may lack some fields. Documents published from now on carry the schema above. Firehose is in preview until 1 December.

We replayed the Firehose from 7 to 13 November over `wss://firehose.nosible.world/v2.0/replay`, with the same API key and the same filters as a live subscription. One request kept the documents that tag Nvidia or Jensen Huang:

// replay request (you send)

Expand all Collapse all Copy

{

type, start, end, filters

}

"type": "replay",

"start": "2026-11-07T11:00:00Z",

"end": "2026-11-13T10:00:00Z",

"filters"

:

{

entities

}

"entities"

:

[

2 items

]

{

label, name

},

"label": "ORG",

"name": "Nvidia"

},

{

label, name

}

"label": "PERSON",

"name": "Jensen Huang"

}

]

}

}

Entities are tagged on English documents, so the chart counts English documents, and we counted each title once per website to drop reposts. The chart below shows the volume of documents, their sentiment against the NVDA share price, and how often four of NOSIBLE's event classes appeared, all on the same hourly axis.

![Stacked panels on one hourly axis for documents tagged with Nvidia or Jensen Huang from 7 to 13 November 2026: documents per hour, sentiment in red with the NVDA share price in white, and hourly counts for disclosed corporate secrets leak, chief executive officer departed, alleged sexual harassment and share price decreased](https://nosible.com/images/2026/11/firehose-nvidia-mix.svg)

Volume stepped up in stages. It went from about 10 documents an hour to 73, 188 and 218 in the first three hours of the story on Sunday evening, settled at a few hundred an hour, and reached 1,561 in the hour from 14:00 UTC on Thursday 12 November, half an hour before the NYSE opened. Markets were closed when the spike began, so the first NVDA price came 13 hours later, at 09:00 UTC on Monday in pre-market trading, 4.9% below its last price on Friday 6 November.

Over the week NVDA fell 9.2% and touched $150.91 at 14:20 UTC on Thursday, ten minutes before the open. Where the chart shows no share price, trading was closed, and a dotted line joins the last price of one session to the first of the next. Sentiment sits in the second panel. It is the probability a document is positive minus the probability it is negative, so anything below the dotted neutral line is negative.

The archive starts on Saturday morning, when the story was already in the news, and the average across all 38,052 documents was -0.38. Sentiment had eased to +0.12 by 13:00 UTC on Wednesday, then fell to -0.15 by 16:00 UTC while NVDA held near $176. NVDA only dropped after the close, from $173.66 at 21:00 UTC to $159.37 at 23:00 UTC, and sentiment reached -0.34 and then -0.50 in those two hours. The four panels at the bottom show what the conversation was about.

Alleged sexual harassment was the biggest class, 12% of all documents, and peaked at 178 documents an hour at 22:00 UTC on Wednesday, midway through the after-hours drop. Chief executive officer departed was 6% of documents but arrived late, peaking at 400 documents an hour at 14:00 UTC on Thursday, the overall peak hour, when NVDA touched $150.91 at 14:20 UTC, closing at $166.00. Share price decreased was 7% of documents, and disclosed corporate secrets leak was 3%.

We have no class for "CEO had an affair", so the model chose the closest one, alleged sexual harassment. Documents that did not carry a sentiment signal when they arrived (about half of them, from before sentiment went live on the stream and during a short outage) were scored afterwards with the same model. Documents now carry their event classes, sentiment and forward-looking classification from the stream, so a client subscribed to Nvidia and Jensen Huang can build everything in the chart except the share price from the stream alone.

## Firehose is Best for Mid-Frequency Strategies

Firehose is best for mid-frequency strategies: those that make tens or hundreds of trading decisions every day across a wide variety of capital markets. Mid-frequency here means between tick data and end-of-day data. Their problem is deciding which of the day's published documents matter to the positions they hold.

Teams use [Atlas](https://nosible.com/blog/announcing-atlas-the-point-in-time-knowledge-graph), World and Search to build their priors, meaning what they already know: how companies connect, what has happened before, and what the evidence says. Firehose makes those priors actionable in real time as stories develop.

A team subscribes to the names, topics and places behind each of its priors. Each matching document arrives as soon as it is processed, already enriched at every level: tickers in the header for routing, and on every snippet the entities, locations, ontologies, sentiment, forward-looking classification and embedding, so a team can act on the exact passage that matters instead of re-reading the page.

## Get Early Access to the Firehose in Preview

Firehose is in preview until 1 December, while we test it in a live production setting with real traffic and fix any issues that turn up. When the preview ends, Firehose will become generally available.

Firehose access is reserved. Each API key is approved individually, and a key that has not been approved is queued and disconnected with `firehose_access_pending`. To request access, email [stuart@nosible.com](mailto:stuart@nosible.com) and tell us which documents your team needs. We will say whether the filters can select them. The [Firehose reference](https://docs.nosible.com/endpoints/firehose-connect) documents every frame.

[All Research](https://nosible.com/blog)

Related Research

![Diagram showing World V1.2 forking into World V2.0 and its nine processing layers.](https://nosible.com/images/2026/10/nosible-world-v2-fork-hero-icons.png)

[NOSIBLE World](https://nosible.com/blog/tag/nosible-world)

[Information Retrieval](https://nosible.com/blog/tag/information-retrieval)

[Web Search](https://nosible.com/blog/tag/web-search)

### [World V2.0 is now Generally Available: A Point-in-Time Database of Global Events](https://nosible.com/blog/world-v2-0-generally-available)

2026-10-28

8 min read

![Atlas Graph Harness workflow: entity and relationship grounding, graph investigation, execution checks, reviewer feedback and delivery of accepted results.](https://nosible.com/images/2026/10/hero-graph-harness.png)

[Trading Signals](https://nosible.com/blog/tag/trading-signals)

[Information Retrieval](https://nosible.com/blog/tag/information-retrieval)

[Web Search](https://nosible.com/blog/tag/web-search)

### [Going From Supply Chains to Trading Signals with a Knowledge Graph You Can Own](https://nosible.com/blog/from-supply-chains-to-trading-signals-a-knowledge-graph-you-can-own)

2026-10-19

7 min read

![Dotted NOSIBLE map showing seven oil chokepoint polygons and 5,261 matching World events for September 2026.](https://nosible.com/images/2026/10/world-geography-chokepoints.png)

[NOSIBLE World](https://nosible.com/blog/tag/nosible-world)

[Information Retrieval](https://nosible.com/blog/tag/information-retrieval)

[Web Search](https://nosible.com/blog/tag/web-search)

### [Polygon Queries: Intelligence Across Space and Time](https://nosible.com/blog/polygon-queries-geospatial-temporal-intelligence)

2026-10-16

8 min read

> NOSIBLE Firehose is a live websocket feed of crawled web documents, with entities, tickers, locations, ontologies, sentiment and embeddings on every snippet.

**URL:** https://nosible.com/blog/a-firehose-of-the-web-already-enriched
