Develop & deploy AI with confidence
Log requests to your DB to analyze, evaluate, & monitor. Just two lines of code to get started.
How it works
Add 2 lines of code to get started
Warehouse every model request to your PostgreSQL database. Use logs to analyze, evaluate, and monitor models.
Run analysis and experiments
We store a customizable JSON object so you can granularly monitor usage, calculate cost, run evaluations, and fine-tune models.
Continuous testing and evaluations
Sample logs and test features in production. Version models and configuration, and get alerts when you need to take action.
Backend workflows designed for engineers
"We experiment with LLM models, settings, and optimizations. Velvet made it easy to implement logging and caching. And we're storing training sets to eventually fine-tune our own models.
"Velvet gives us a source of truth for what's happening between the Revo copilot, and the LLMs it orchestrates. We have the data we need to run evaluations, calculate costs, and quickly resolve issues."
"Our engineers use Velvet daily. It monitors AI features in production, even opaque APIs like batch. The caching feature reduces costs significantly. And, we use the logs to observe, test, and fine-tune."
Flexible infrastructure for scale
Data interoperability
Log requests to your database. Secure and compliant.
Granular analysis
Query data with SQL to analyze usage, cost, and metrics.
Intelligent caching
Optimize costs and latency with request caching.
Evaluations framework
Configure evaluations to test your models, settings, and metrics.
Experiments & monitoring
Run evals on subsets of logs, or continuous testing in production.
Dataset generation
Export datasets for fine-tuning models and other batch workflows.
Q & A
Articles from Velvet
Why Find AI logs OpenAI requests with Velvet
AI-powered B2B search engine logged 1,500 requests per second.
Create a fine-tuning dataset for gpt-4o-mini
Use Velvet to identify and export a fine-tuning dataset.
Cache LLM requests to reduce latency and costs
Return results in milliseconds and don't waste calls on identical requests.
AI gateway