Feature Requests

Users can currently install additional libraries in their notebook session environment (via pip install ), but it would be nice to be able to include a requirements.txt or something similar.

Currently notebook sessions simply terminate, relying upon Volumes for persistence between separate sessions. The ability to resume sessions - or perhaps as a mechanism to simply recreate a previous session - would be a nice QoL improvement.

An ability to upload static html reports generated by protocols.

A simple CLI for the BioLM SDK, primarily as an integration point for CLI based workflow tools.

A centralized system for storing and versioning generated sequences, with built-in lineage tracking, annotations, and comparison tools. It provides a single source of truth for sequence records and integrates with external workflows and analysis tools.

An agent that takes a draft notebook or description of a workflow, breaks it into tasks, and produces either polished Jupyter notebooks or formalized pipelines using frameworks like Nextflow, Kedro, or CWL.

Facilitate hosting of open competitions (like the Adaptyv EGFR one) but integrated into Console, allowing users to benchmark their molecules against public datasets.

Users can publish protocols to a community repo, fork others’ protocols, and remix workflows like open-source code.

Ability to fine tune your own models

Automate discovery and extraction of relevant molecular, genomic, and proteomic data. Enable configurable workflows to support diverse dataset needs (e.g., therapeutic proteins, synthetic constructs, IP-free design spaces). Standardize and normalize raw inputs into BioLM’s internal data schema. Provide auditability and metadata tracking for reproducibility and IP clarity. Seamlessly interface with downstream processes (e.g., Fine-Tuning Agent, etc).