Ligant Information

Ligant builds free, open-source tools for target assessment in cell and gene therapy. Each tool answers one narrow question, shows its working, and can be verified independently. They are deterministic, run in the browser or on your own hardware, and collect nothing. No accounts, no telemetry, no analytics. Source under Apache 2.0 at github.com/Ligant-ai. One tool is live. The Antigen Density Calculator converts flow cytometry median fluorescence intensity into antibody binding capacity by calibrating against certified bead standards. The arithmetic is a two-parameter regression that any spreadsheet can do. What the tool adds is a set of diagnostics that catch calibration failures that a goodness-of-fit statistic misses. A standard transcribed from the wrong column fits with a slope of 1.000 and an R² of 1.000, flawless by every conventional criterion, while reporting values roughly four times too low. When a check like that fails, the tool withholds the figure rather than printing a confident number with a caveat underneath. Everyone has a calculator. Almost nobody has a checker. benchtools.ligant.ai. No registration. Click Load worked example and everything above is checkable in under a minute. We got here by being wrong first. In August 2026, we retracted the original product. This single agent produced a finished assessment and published the three findings behind it: proprietary data does not belong in a cloud AI, the system handed people a conclusion rather than the reasoning, and the target assessment is not a linear process. For regulated organizations, we deploy privately on your infrastructure, with connected workflows that carry values and their flags between tools, local language models, and the validation evidence and package your quality unit executes against. The free tools are not a trial of it. hello@ligant.ai