Python client for Engrammatic — ingest-time anomaly, novelty and fraud detection with binary hyperdimensional computing.
Project description
Engrammatic Python client
Ingest-time anomaly, novelty and fraud detection with binary hyperdimensional computing — encode each event once, compare with XOR + popcount. This package is the official Python client for the Engrammatic service. It has zero dependencies (stdlib only).
Install
pip install engrammatic
Quickstart
Sign in with Google at app.engrammatic.ai — new accounts get $5 of starting credit — and copy your API key from the dashboard.
from engrammatic import Client
c = Client(api_key="ek_...")
# score one event (and learn it)
r = c.detect({"path": "/wp-admin.php", "status": 404, "ua": "python-requests"})
print(r) # {"score": ..., "best_similarity": ..., "is_anomaly": ..., "balance_usd": ...}
# batch log ingestion
c.ingest([{"path": "/api/v1/orders", "status": 200, "ms": 41},
{"path": "/api/v1/orders", "status": 503, "ms": 9040}])
# similarity search over what the memory has seen
hits = c.query({"path": "/api/v1/orders", "status": 503}, k=5)
# store without scoring
c.write({"path": "/healthz", "status": 200})
# current plans/rates (no key needed)
print(Client().pricing())
Out-of-credit responses raise EngrammaticError with .out_of_credit == True
(HTTP 402) — top up in the console.
What it's for (and honestly, what it's not)
Engrammatic sits alongside your log platform (Hydrolix / Splunk / Datadog), scoring events at ingest time for combinatorial novelty — field combinations your stack hasn't seen — at $0.30–$1.00 per million events. If your problem is a 5xx spike, use a threshold; a numeric outlier, a z-score; known signatures, a WAF. Engrammatic earns its keep on the anomalies those baselines can't express.
Notes
- This release is the remote client only; the embedded/local engine is not
included (
Client(local=True)raises). - AI agents: the service publishes a manifest at
https://app.engrammatic.ai/.well-known/agent.jsonand supports programmatic top-up — see the console. - Docs & contact: engrammatic.ai · hello@engrammatic.ai
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file engrammatic-0.1.0.tar.gz.
File metadata
- Download URL: engrammatic-0.1.0.tar.gz
- Upload date:
- Size: 4.7 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.12.13
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
6b1dc6769b5c7e466ccdec70ba6e52b358033ab57264df7af954a2757342ddb6
|
|
| MD5 |
bcfab8d7c2073ce294be888d40fe8f5c
|
|
| BLAKE2b-256 |
5b179d18a5fdc281b04293db31d5ff29db2b272bfde4ab69549d268fd1ff0b36
|
File details
Details for the file engrammatic-0.1.0-py3-none-any.whl.
File metadata
- Download URL: engrammatic-0.1.0-py3-none-any.whl
- Upload date:
- Size: 6.1 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.12.13
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
39f942e128d46ca2091d654003de7de882472f9319fda3c3a42d292ed900c6c9
|
|
| MD5 |
450191d811a9cf9ddfbbb90bfcef4e14
|
|
| BLAKE2b-256 |
67dcc03d87e07344caa5fbbfaf575c53f1ba96a6f402a967d2934580f6795380
|