Hi Parakeet AI Team,
I would like to report two issues I am facing while using Parakeet AI for technical interview preparation.
First, I noticed that the knowledge cutoff is showing as June 2024. Because of this, some answers are outdated and do not reflect the latest information, features, or practices related to current technologies and performance testing tools. It would be helpful if the knowledge base could be updated with more recent information.
The second issue is with the way interview answers are generated. I need answers that sound natural and human-like, as if an experienced professional is actually answering the interviewer.
For example, my resume contains multiple performance testing tools such as Dynatrace, AppDynamics, Grafana, Jenkins, AWS, and Python. If the interviewer asks a question specifically about JMeter, the answer should mainly focus on JMeter and the context of the question. However, Parakeet AI sometimes tries to include almost all the tools mentioned in my resume, even when they are not relevant to the question.
This makes the response sound too scripted and AI-generated, rather than like a natural interview answer.
I would prefer the answers to focus only on the tool or technology relevant to the question, without unnecessarily mentioning every technology from my resume. The responses should sound conversational and natural when spoken, while remaining concise and technically strong. Where appropriate, they should include realistic project-based examples, avoid repetitive or overly formal textbook-style wording, and use the latest available technical information.
Could you please look into both the knowledge cutoff and the answer-generation style and improve them?
These improvements would make Parakeet AI much more useful for real-world technical interview preparation.