QSIMpost‑Opt is an AI‑applied software library that automates the post‑processing step of Quantum Key Distribution (QKD). In QKD, encryption keys travel as single photons; if anyone tries to snoop, the photons change and the intrusion is exposed. Converting this raw signal into a final cryptographic key normally means weeks of expert tuning (decoy ratios, detector‑noise thresholds, protocol settings, or sharing sensitive design data with outside consultants). QSIMpost‑Opt plugs into existing QKD hardware, learns device behavior in real time, and uses machine learning to apply the best parameters for each installation and changing environment. It supports multiple QKD protocols, runs local attack simulations, and delivers the longest safe key the system can achieve. Customers cut development time and cost by 75% while keeping all configuration data in‑house.