General Questions
General Questions
1. What is HERO?
HERO is a virtual coach powered by artificial intelligence and part of the Happyforce platform. Its goal is to provide personalized support to employees through meaningful conversations and practical tools that promote well-being, engagement, and happiness at work.
2. How does HERO work?
HERO interacts in two main ways:
Personalized Conversations: Engages users to explore emotions, challenges, or reflections in a supportive manner.
Practical Insights: Provides brief, actionable advice to inspire small, positive changes.
HERO combines the user's personal context and organizational priorities to deliver a unique and relevant experience.
3. Is HERO a chatbot?
No, HERO is much more than a chatbot. It is a hyper-personalized coach that combines empathy with advanced artificial intelligence to provide practical guidance tailored to individual needs, rather than relying on predefined responses.
4. Are user data safe with HERO?
Yes, data security and privacy are top priorities:
Complete Confidentiality: Interactions with HERO are private and are not shared with others in the organization.
Anonymized Data: Reports or analyses generated from interactions are presented in an aggregated format, ensuring no user can be identified.
Regulatory Compliance: HERO adheres to global data protection regulations such as GDPR.
Data Control: Users can request the deletion of their information at any time.
5. Who can use HERO?
HERO is available to employees of organizations that have integrated Happyforce and HERO into their workplace well-being initiatives. To function properly, the email registered with Happyforce must match the one used within the organization.
6. How does HERO personalize its support?
HERO tailors its content and guidance by combining:
User's Personal Context: Mood, well-being trends, and feedback collected in Happyforce.
Organizational Context: Strategic goals, cultural values, and specific company metrics.
This ensures that HERO provides practical recommendations aligned with the user’s needs and the organization’s objectives.
7. What type of content does HERO offer?
HERO provides:
Practical Guidance on topics such as well-being, emotional management, and work-life balance.
Motivational Reflections that foster engagement and personal development.
Content Aligned with the organization's strategic priorities and values.
8. How often does HERO interact with users?
HERO sends messages and ideas at regular intervals designed to be light and non-disruptive to the user’s work activities.
9. Can managers access interactions with HERO?
No. Managers do not have access to individual conversations between HERO and users. They only receive anonymized and aggregated data to support the overall well-being of their teams.
10. Is using HERO mandatory?
Using HERO is entirely optional. Users can choose how and when to interact with HERO and, if preferred, may request its deactivation through the Happyforce administrator.
11. Does HERO work offline?
No, HERO requires an active connection to Happyforce to provide its personalized support.
12. How can feedback about HERO be shared?
Feedback or suggestions about HERO can be shared directly through the Happyforce platform under the help section. Your feedback is always welcome!
Security & Privacy
Security & Privacy
1. What happens to the data collected by HERO?
Regarding data usage and protection:
The data from conversations and interactions with employees is used exclusively to personalize the user experience and improve service quality.
Data is not shared with third parties under any circumstances.
All information processing is performed in an aggregated and anonymous manner.
Any data that could individually identify employees is deleted.
This approach allows us to maintain a balance between HERO's personalization and the protection of our users' privacy.
2. How is data privacy and anonymity ensured?
All interactions with the language models are conducted anonymously and securely. Before any interaction, we implement a rigorous process to remove sensitive information using advanced natural language processing techniques. This ensures that no personal or confidential data is transmitted during conversations.
The system employs multiple layers of protection to safeguard user privacy. In addition to anonymization, we use data masking techniques and sensitive content filters to maximize security.
The models we use strictly adhere to not training on data sent to their APIs, providing additional assurance that shared information will not be used for future training.
3. What measures are in place to prevent data leaks?
Happyforce implements robust measures to prevent data breaches, including:
Data encryption in transit and at rest.
Role-based access controls (RBAC) and multi-factor authentication (MFA).
Anonymization of data to ensure privacy and protection against breaches.
Continuous monitoring using tools like Sentry and Papertrail.
A secure development lifecycle (SDLC) with regular security reviews and testing.
Additionally, Happyforce evaluates external vendors, provides regular security training to its employees, and maintains an incident response plan to mitigate any risks. These actions, aligned with regulations like GDPR, ensure comprehensive information protection.
4. How does HERO prevent bias in its generated responses?
Through the combination of advanced prompting techniques and the latest large language models (LLMs), we ensure safe and high-quality interactions.
Users can report any inappropriate interactions, enabling continuous system improvement.
Interactions are designed to align with the values of each organization, limiting inappropriate responses.
This advanced prompting system and powerful LLMs enable careful and secure interactions. User feedback acts as a safeguard, allowing dynamic adjustments to the model while maintaining necessary oversight.
5. What does HERO do to ensure ethical operation?
The techniques mentioned above also apply here. Additionally, the developers of the advanced models used by Happyforce have implemented specific safeguards to ensure the ethical behavior of the models.