The obesity epidemic is one of today's most notable public health problems. Like many physiological parameters and lifestyle habits, the glucose metabolism and eating style varies among people not only by average amounts of consumed calories and macronutrients, but also by their distribution throughout the day — early or late eating, eating more and less frequently, etc. Research study results reported in scientific literature indicate that on average, late eating is associated with higher obesity risk and lower efficiency of weight loss treatment. More rigorous insights require proper account of personal behaviours such as chronotype and its stability. Lack of unique data format and processing techniques leads to inconsistent approaches and mixed results, hard to combine even in meta studies. We offer a tool for wearables data processing and circadian feature extraction that can solve this problem and provide more robust results with reference to personal circadian traits and habits, as several works suggest: