Production infrastructure for HyperCycle AIMs — telemetry, retries, health, and inter-AIM routing.
Project description
qyra
Python SDK for the Qyra AI infrastructure layer.
Qyra runs production AI Modules (AIMs) on the HyperCycle network — covering scoring, routing, research, document intelligence, personal memory, benchmarking, telemetry, and certification. This SDK lets you call any AIM, track costs, and log telemetry from your own Python code or agent pipelines.
Installation
pip install qyra
Quick start
from qyra import QyraClient
client = QyraClient(base_url="https://api.qyratech.com")
# Call an AIM
result = client.call("aim-smart-router", {"query": "Summarise this document", "payload": "..."})
print(result)
# Track a call with telemetry
client.track(
aim_id="aim-smart-router",
latency_ms=142,
cost_usd=0.0028,
status="success"
)
Telemetry
The SDK ships with a built-in telemetry client that posts structured call data to qyra-ingest. Every call can log latency, cost, model used, and outcome — visible in real time on the Qyra dashboard.
from qyra.telemetry import TelemetryClient
tel = TelemetryClient(ingest_url="https://api.qyratech.com/ingest")
tel.track(aim_id="aim-scoring", latency_ms=88, cost_usd=0.0086, status="success")
Production AIMs
| AIM | Description | Cost/call |
|---|---|---|
aim-scoring |
Survival scoring and ΩSCORE modelling | $0.0086 |
aim-smart-router |
Intelligent multi-AIM routing | $0.0028 |
aim-personal-memory |
Persistent memory across sessions | $0.0015 |
aim-web-research |
Autonomous web research | — |
aim-smart-document |
Document intelligence and extraction | — |
aim-benchmarker-v2 |
Empirical cost/quality benchmarking | — |
aim-certification |
AIM quality certification | — |
aim-ingest |
Telemetry ingest service | — |
Links
- Website: qyratech.com
- Book a call: cal.eu/qyratech/discovery
- GitHub: github.com/qyrahub
Changelog
v0.1.5
- Corrected telemetry ingest URL
- Server-side UUID fallback for older clients
- Real cost derivation in aggregate endpoint
v0.1.2
- Fixed telemetry URL pointing to production ingest
- P50/P95/P99 latency percentiles in dashboard
v0.1.0
- Initial release
QyraClientwith AIM call support- Telemetry client with structured logging
Built in South Africa. Infrastructure for the agentic age.
Project details
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 qyra-0.1.6.tar.gz.
File metadata
- Download URL: qyra-0.1.6.tar.gz
- Upload date:
- Size: 24.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.12.3
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
34a367878dcba18e2ae087a1063bfbabd9fe80a51ed4530e8ddb5bd839a74711
|
|
| MD5 |
3fbfcce66ec2bd9dd20c792bb52914e0
|
|
| BLAKE2b-256 |
46df4899cad776f97242b04e49a7f97f2620b911544efd9b48adf2907e499a8d
|
File details
Details for the file qyra-0.1.6-py3-none-any.whl.
File metadata
- Download URL: qyra-0.1.6-py3-none-any.whl
- Upload date:
- Size: 20.9 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.12.3
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
885208774860e7e7a149a84d3a6ab699ca66965f2855e7a30a1d8dff69dc9165
|
|
| MD5 |
422bcfc8f3ca3112bf937d86e49e09bf
|
|
| BLAKE2b-256 |
f0aa1597596f3cb4bf0f429bd52275c1f28fb6c133e8d35b552febfaaa293f7c
|