Roadmap

Feel free to email us here to request features or to request them sooner. We push updates weekly and will make changes available in our changelog

​

October

• Automatically version models with every deploy
• Allow users to see API logs
• Rollback models to previous versions with one click
• Implement serverless based GPU’s for larger model pipelines

​

November

• Implement post-processing functions
• Implement LLM’s (SD2, Whisper, FlanT5 etc)
• Infrastructure improvements (cold-start, latency)
• One-click deploy models

​

December

• Model monitoring integrations
• Improve LLM cold-start times
• SOC 2 Type 1

​

January

• Redid internal architecture
• Increased inference times across all pre-built models

​

February

• Add HuggingFace integration
• Add Spacy support
• Add more pre-built models
• Add better logging

​

March

• Make fine-tuning publicly available in Alpha
• Allow users to deploy any custom code (Alpha)

​

April

• Make fine-tuning publicly available in Beta
• Make custom code functionality available in Beta
• Make multiple GPU’s available