(future) Collect results from runs, and store in a db

[willcl-ark]

After a benchmark is complete, we should zip and ship results to an s3 (or similar), and process it into a DB.

Required work first:

• Benchmarks to be run
• Benchmark results to be pre-processed