Safety and Feasibility of a Machine-Learning Bolus Priming Added to Existing Control Algorithm

This study is testing a new machine-learning feature that can be added to an existing automated insulin delivery (AID) system to improve mealtime insulin dosing for people with type 1 diabetes. The goal is to make blood sugar control around meals safer and more effective. **You may be eligible if...** - You are 18 or older with type 1 diabetes for at least one year - You have used an automated insulin delivery system with a Dexcom G6 or G7 continuous glucose monitor (CGM) in the past 3 months - You have been using insulin for at least 6 months - You have someone nearby who knows how to handle a low blood sugar emergency - You are not pregnant or breastfeeding - You speak and read English - You have internet access at home and are willing to upload data as needed **You may NOT be eligible if...** - You are planning to start a new non-insulin glucose-lowering medication - You currently use an SGLT-2 inhibitor (like empagliflozin or dapagliflozin) - You have had a seizure or lost consciousness from low blood sugar in the past 12 months - You have had a diabetic ketoacidosis episode in the past 12 months - You have advanced kidney disease or are on dialysis - You are being treated for a seizure disorder or adrenal insufficiency - You are currently in another interventional trial Talk to your doctor to see if this trial is right for you.