Hilly AI
Expanding access to care with clinical-grade conversational AI
A conversational AI assistant designed to support clinical triage, medical assessment, and follow-up in high-demand healthcare settings, that conducts structured assessments aligned with validated clinical pathways and always under human supervision.
Patient communication received
Next appointment prep
Today
Communication sent
Today
3/3 received
Today
Automations
Patient response details
Next appointment preparation
Ongoing
3/3 received
Next communication
Tomorrow
Enter your current weight
Are your ankles or legs more swollen than normal?
Clothes or shoes tighter than normal?
Have you been sleeping with more pillows than normal?
Trusted by Clinical Teams
+1 Million
Hilly-mediated clinical interactions70%
High patient engagement with weekly follow-up communicationsx2
Follow-up capacity
Where AI Voice Meets Clinical Precision
Turning patient conversations into consistent, clinically validated decisions.
Hilly AI brings together natural conversational interaction with clinically validated logic, ensuring safe, explainable, and reproducible patient assessments. The system combines advanced speech-to-text, natural language understanding, and text-to-speech modules with the certified Route engine, which executes validated clinical pathways.
Fully human-supervised, Hilly AI enables healthcare services to scale remote patient assessment and follow-up efficiently.
UpHill is compliant with relevant industry-standard certifications
/f/152598/2400x1706/cc7d3945e4/marca_ce.png)
/f/152598/1500x850/fd10b5c6e3/is-788769.png)
/f/152598/1200x1297/1480bd46e5/iso-13485.webp)
/f/152598/143x142/17d47a27e9/gdpr.png "GDPR")
/f/152598/579x350/ff390abd46/ens.png)
/f/152598/1182x1182/63ff0a3c4a/selo_id_reconhecimento_idoneidade_cores.png)
Multiple innovations, a single purpose: better care
Keep patients connected between visits
Extend clinician–patient communication beyond the hospital, enabling early identification of changes in condition and timely intervention before escalation is required.
Prepare clinical interactions before they begin
Prepare clinical consultations in advance by collecting and structuring patient information before the interaction, enabling faster, more focused, and more effective care.
Turn ongoing interactions into actionable care updates
Automate the collection of clinical information to track patient progress, surface relevant changes, and support timely clinical decisions and follow-up actions.
Give healthcare professionals time to focus on what matters: the patient.
Hilly AI takes care of repetitive, time-consuming tasks, allowing healthcare professionals to spend more time on what matters most: patient care.
Enabling Safe, Scalable, and Intelligent Care
Explainable and Auditable Decision-Making
All clinical recommendations are fully traceable and transparent, providing reproducible outputs that can be reviewed, audited, and verified.
Multilingual Support & Natural Conversation
Delivers natural, multilingual interactions that make healthcare more accessible and inclusive, allowing patients from diverse linguistic backgrounds to engage seamlessly.
Safe and Compliant with EU AI & MDR Regulations
Operates as a high-risk, EU-regulated AI system integrated with a Class IIa medical device, ensuring adherence to MDR 2017/745 and EU AI Act requirements for transparency, safety, and clinical oversight.
Human-in-the-Loop Supervision
Every output from the system is reviewed, verified, and validated by a qualified nurse before any clinical action, maintaining safety, accuracy, and professional accountability.
Multi-channel patient access
Engages patients seamlessly across multiple channels - webforms, kiosks, web chat AI, voice AI, automated calls, and email - allowing care teams to reach patients where they are and capture information efficiently.
Rigorous Development & Validation Lifecycle
Tracking accuracy from day one
Automated metrics such as Word Error Rate (WER) are applied from early stages to continuously track model performance before human validation.
Automated evaluations guide model development across its entire lifecycle, supporting informed, evidence-based decisions.
A dedicated comparison dataset ensures the model evolves according to real-world safety, accuracy, and clinical relevance benchmarks.
Clinical rigor you can trust
A validation platform compares system outputs against gold-standard, clinically validated transcripts. Clinical reviewers assess results in a blinded setup to ensure scientific rigor and impartiality.
The process includes regression testing, edge-case testing (e.g., ambiguous language), and mandatory safety checks using critical scenarios.
Ensuring fair and reliable AI for all
Performance is assessed across diverse patient profiles—accents, regions, and age groups—ensuring non-discrimination and alignment with ethical AI and EU AI Act requirements.
Continuous audits and predefined evaluation criteria, based on real clinical cases and simulated scenarios, ensure consistent quality across versions.
More of UpHill
UpHill provides solutions for companies and professionals across the healthcare industry.
