AI-driven trading
strategies, onchain.
A portfolio of fully automated strategies trading single-name equities, ETFs, commodities and crypto perpetuals on Hyperliquid.
Backtested over a 12-month window (Mar 2025 – Mar 2026). Past performance is not indicative of future results. Live performance and full results are published at automate.precipitate.ai.
Chat with the same intel
our strategies trade on.
Oracle is an AI research assistant grounded in our daily Hyperliquid whale-positioning briefs. Ask about specific assets, positioning shifts, or today's trade suggestion — gated to RAIN holders.
0xb83de01…d6e36 carries $22.1M of it with +$9.6M unrealized profit and is not flinching.Production-grade compute
for AI workloads.
NVIDIA A16 / A40 / A100 GPU servers, open-source LLM endpoints, and CPU servers tuned for high-frequency automation. Hosted in Mumbai for low-latency execution against onchain venues.
Cloud Claw
Optimized for automation & high-frequency trading workloads.
- 6 vCPU
- 12 GB DDR4
- Mumbai · Low Latency
NVIDIA A16
AI inference, virtual workstations, GPU-accelerated apps.
- 2 vCPU
- 16 GB RAM
- NVIDIA A16
NVIDIA A40
AI/ML training, 3D rendering, high-performance compute.
- 6 vCPU
- 30 GB RAM
- NVIDIA A40
NVIDIA A100
Enterprise-grade AI/ML workloads with 24×7 support.
- 6 vCPU
- 60 GB RAM
- NVIDIA A100
GPT-OSS 120B
OpenAI open-weights LLM with chain-of-thought, coding & tool use.
- 6 vCPU
- 80 GB RAM
- NVIDIA A100
DeepSeek 32B
Coder + LLM endpoints for code generation and reasoning.
- 24 vCPU
- 120 GB RAM
- Coder + Chat