Startup funding data analyzer
- Set up a database instance with the schema in the schema.sql file.
- Use the import command (i.e., see
/src/api/import.rs
) to ingest funding data into the db instance. (Feel free to ingest csv files in the /data directory) - (Optional) If you have an account on Crunchbase, download the data in csv format by clicking "EXPORT TO CSV" here.
- Querying pre-seed & seed startup funding data based on the following criteria: "Industry", "Last X days", & "Currency".
- Importing a csv file into the db.
- Asking questions for the fetched funding data in natural language.
cargo build --release
- Import:
./target/release/crunchbase import
- Query:
./target/release/crunchbase query
- Set up a central database for others to use upon approval.
- Integrate an LLM-based agent with tools (e.g., search, access to vector store, etc.)
- Add prompt templates
- Add plot capability
- The quality of the answer is almost always better with GPT-4, but be aware of the cost.
- If the context length limit is reached based on the data points, consider modifying the number of days to resolve the problem.