Disertus Labs (Milo AI Speech Therapist)

Disertus Labs’ Milo AI Speech Therapist occupies a narrow but meaningful niche: it is a 24/7 conversational agent designed to help users practice fluency, articulation, and clarity outside the confines of a traditional therapist’s office. Accessible via iMessage, Web, and WhatsApp, Milo is not a replacement for clinical speech therapy, but rather a supplementary tool that fills the gap between scheduled sessions. Its core value proposition rests on three pillars: real-time conversation with instant feedback, post-session summaries with speech metrics, and a library of 47 guided exercises across 12 categories. For users who need a low-pressure environment to rehearse interviews, presentations, or daily conversations, Milo offers a practical, always-available sandbox. The live transcription feature automatically detects filler words and repetitions, providing immediate, actionable cues that can help users build awareness of their speech patterns. After each session, Milo generates a summary that includes metrics like speaking rate and filler word frequency, along with personalized tips and exercise recommendations. This combination of real-time correction and reflective analysis is where Milo differentiates itself from generic voice assistants or recording tools. However, the tool’s effectiveness is constrained by several factors. The free Starter plan limits users to 20 voice minutes per month and only three research chats, which is sufficient for occasional practice but inadequate for consistent daily use. The Pro plan, at $20 per month (or $16 annually), removes those caps and unlocks the full exercise library and advanced tracking, making it a more realistic option for committed users. Yet even at that tier, there is no mention of licensed therapist oversight or clinical validation, which raises questions about the accuracy of feedback and the safety of relying on AI for speech concerns that may require professional diagnosis. Data encryption is stated, but compliance details such as HIPAA are not disclosed, which may be a concern for users handling sensitive health data. The exercise library covers breathing, pronunciation, fluency, and other categories, but the depth and adaptability of these exercises remain untested in this review. For users with specific disorders like cluttering or apraxia, the available exercises may not be tailored enough, and Milo’s feedback is likely optimized for common issues like stuttering and filler word overuse. The most practical use cases are daily fluency practice for stuttering, interview preparation, public speaking rehearsal, and supplementing traditional therapy. In each scenario, Milo serves as a judgment-free practice partner that can be used on demand, but its limitations in clinical depth and personalization mean it should be viewed as a complement to, not a substitute for, professional guidance. Ultimately, Milo is best suited for motivated individuals who want to extend their practice hours and gain quantitative insight into their speech patterns, provided they temper expectations around clinical rigor and data privacy.