What it does
It offers the opportunity to ask general questions about a log file instead of searching for specific strings only.
How we built it
It takes a log file as an input, then the data is split into chunks. After that we create a Chroma Vector Database and query the vector database with Langchain.
Challenges we ran into
Combining the knowledge of a LLM with a decent interface was quite challenging, but in the end it almost feels like ChatGPT.
Accomplishments that we're proud of
Dora is able to find details in the log file we couldn't find on our own. It also always shows the original data its answer is based on.
What's next for Dora the Logexplorer
The UX always needs improvement. To deploy it into an application available on the market, it should be tested with a much larger set of potential log files.
Built With
- llm
- python
Log in or sign up for Devpost to join the conversation.