Defined in Python. Computed in Rust.

Every feature, right on
thyme.

Thyme is the streaming feature platform that eliminates training/serving skew. Define features once in Python — Thyme compiles them to a high-throughput Rust engine for real-time serving and point-in-time correct offline training.

P99 Latency
<5ms
🕐
Feature Freshness
Real-time
Training/Serving
Zero Skew
Features

Everything your ML pipeline needs

From feature computation to serving, Thyme handles the entire lifecycle so your team can focus on building great models.

Rust-Powered Engine

Features defined in Python are compiled to a high-throughput Rust streaming engine. Real-time aggregations with millisecond freshness.

Time-Travel Queries

Point-in-time correct feature retrieval for training. Query any feature exactly as it was known at any past moment.

Zero Training/Serving Skew

One definition, two modes. The same feature logic runs in both streaming aggregation and offline point-in-time lookups — no divergence, no silent accuracy drops.

Datasets, Pipelines & Extractors

Composable abstractions: datasets define event streams, pipelines apply windowed aggregations, and extractors compute derived features on read.

Exactly-Once Semantics

Distributed leasing, checkpointing, and replay logs ensure exactly-once processing with no data loss or duplication.

Declarative, Not Operational

No Kafka consumers to manage, no state stores to tune, no checkpoint recovery to handle. You own the feature logic — Thyme owns the infrastructure.

Architecture

Built for the modern ML stack

Define features in Python, commit them with the CLI, and Thyme compiles them through a four-layer architecture — from SDK to serving — all powered by Rust.

Python SDK
DatasetsPipelinesFeaturesetsExtractorsSources
Control Plane
Definition ServiceGraph ValidationBlueprint PlanningJob Scheduling
Data Plane
Rust Streaming EngineWindowed AggregationsRocksDB StateKafka Transport
Query Layer
Online ServingPoint-in-Time LookupsExtractor DAGFeature API
Powered by Thyme
Performance

Built for simplicity and speed

Thyme compiles Python feature definitions to a Rust streaming engine. Low latency, zero skew, and a three-command deployment workflow.

<0ms

P99 Online Latency

0

Definition for Online & Offline

0

Training/Serving Skew

0

Commands to Deploy

features.py
from thyme import *

@dataset(index=True)
class UserStats:
    user_id: str = field(key=True)
    ts: datetime = field(timestamp=True)
    avg_spend_7d: float

    @pipeline(version=1)
    @inputs(Transaction)
    def compute(cls, t):
        return t.groupby("user_id").aggregate(
            avg_spend_7d=Avg(of="amount", window="7d")
        )

Define features in Python. Deploy with thyme commit. Serve in milliseconds.

It's about thyme you upgraded
your feature platform

Join the teams shipping ML features faster with Thyme. Get up and running in minutes, not months.