Timber
Ollama for classical ML models, 336x faster than Python
Get picks like this daily. Today's launch numbers, the top launches, AI and tech news, and a daily opportunity pick, straight to your inbox.
Want to reach these readers? Sponsor an issue.
HN · launched Mar 2, 2026 · 207 upvotes (top 3%) · 0 views
Niche: machine learning inference engines · 108 launches, 8 in the last 7 days ▼ -27% · Explore this niche →
Category and tags
- Horizontal
- Model & infra
- Developer
- AI feature
- Hobby / open-source project
Visit website ↗ Source listing ↗
More like this: all 108 launches in machine learning inference engines →
Could you build this?
No Re-implementing a high-performance C++ or Rust inference engine for classical ML models that is hundreds of times faster than Python requires specialized systems and SIMD optimization skills.
What it would actually take: Building an Ollama equivalent for classical ML requires writing a native binary runtime (in Rust, C++, or Zig) that parses serialized model weights (GBDTs like XGBoost/LightGBM, Random Forests, Linear models) into cache-aligned decision trees with SIMD/AVX vectorization and concurrent thread pools, accompanied by a model registry and unified REST/gRPC API. This requires deep low-level systems engineering, cache-locality optimization, and knowledge of classical ML internal data structures.
Discussion
20 comments analyzed.
Competitors mentioned: ONNX, xnnpack, lightgbm, llama-cpp, vLLM
Concerns raised: Swapping backends in production can be far from trivial, Feature extraction/transformation step is often the bottleneck, not inference, Only useful if you already have solved data plumbing and optimized pipeline, C compiler toolchain not available on most Unix systems, Requires pre-formed feature vectors as input
Feature requests: Accept raw input (before feature extraction) instead of just feature vectors, Add performance comparisons vs vanilla inference latencies, Support for numeric data feature extraction with LLM
Competitors
Other products that read as similar to this one : 2617 launches clear the similarity bar, closest 8 shown.
Attention rank: #104 of 2618 (itself plus its competitors, highest first, normalized so YC and Product Hunt are compared fairly).
Launched 124 days after the earliest competitor.
- JevGym · github · 2026-09-22 · 27 upvotes · similarity 0.60
- EVA · github · 2026-10-06 · 10 upvotes · similarity 0.59
- OpenLake: Fast inference for self improving models · yc · 2026-11-05 · 63 upvotes · similarity 0.58
- Goku · hn · 2026-07-15 · 9 upvotes · similarity 0.57
- OM Core · hn · 2026-06-30 · 14 upvotes · similarity 0.57
- zero2vllm · github · 2026-10-12 · 14 upvotes · similarity 0.57
- vllm-jev · github · 2026-09-24 · 72 upvotes · similarity 0.56
- genpark-dense-feedforward-mlp-backprop-skill · github · 2026-09-28 · 7 upvotes · similarity 0.56
The niche, labels, vibe code rating and competitor comparison on this page are analysis by Nicheloom. If you use them, please credit nicheloom.com. How you may use our data.