1. What our AI does
AI powers CV review scoring and suggestions, interview practice questions and evaluation, and an in-app assistant. It does not make hiring decisions and does not send communications on your behalf.
2. Models and providers
We route model calls through the Lovable AI Gateway to modern large language and speech models. Providers are contractually bound not to train foundation models on your inputs.
3. Limitations
- AI can be confidently wrong ("hallucinate"). Verify important facts.
- AI may miss context specific to your role, sector or country.
- Output quality depends on the quality of the input you provide.
- Speech-to-text can misrecognise names, acronyms and accents.
4. Human oversight
Coaches review flagged outputs; users can escalate any AI report to a human coach. Our team monitors evaluations for regressions and safety issues.
5. Bias and fairness
LLMs are trained on internet-scale data and can reflect societal biases. We regularly test for demographic bias in scoring and adjust prompts and calibration accordingly. If you believe an output was biased, please tell us.
6. No employment guarantee
Using NEguso does not guarantee interviews, offers or salary outcomes. Career outcomes depend on many factors outside our control.
7. Not legal, medical or financial advice
AI output is informational only. It is not legal, medical, financial or immigration advice. Consult qualified professionals for decisions in those areas.
8. Your responsibility
- Review AI output before sharing it externally.
- Never submit fabricated experience to employers.
- AI is not a substitute for professional legal, medical, financial, immigration or other regulated advice.
- Human review is available for applicable coaching services — you can escalate any AI report to a human coach.
- Report concerns or challenge an AI output by emailing privacy@negusoaicareerhub.ie.
9. EU AI Act alignment
NEguso's AI features are limited-risk under the EU AI Act. We disclose AI involvement in interactions, label AI-generated content, and provide meaningful information about the logic involved on request.