How do you build the prefixes when multiple words share the same prefix?
If my understanding is correct, the method is:
1. Find common search terms
2. For each search term (eg test, compute it's vector [1.23, 4.56...] and it's prefixes [t, te, tes...])
3. Store these in Qdrant as t->[1.23, 4.56...] , te->[1.23, 4.56...] , tes->[1.23, 4.56...] and so on. Here, each of the prefixes are used as point_ids
4. When a search query comes in, call /recommend and pass in the partial query as the point id
Hey Nikita! I was just looking at the docs but I was a bit confused about what the various compute instances were doing. Do they all serve reads and writes? If so, is there data partitioning or does this support distributed transactions?
Various compute instances are different endpoints to separate databases. So for now it's single writer system. You can get a lot of power out of a 128 core compute node. In the future will will also spin up extra compute to scale reads.
In the future after that future we will introduce data partitioning - we have a cool design for it, but one step at a time.
Notion | Head of Platform Engineering & Head of Product Engineering | San Francisco, CA
Here at Notion, our work shapes our culture and our culture inspires our work. We seek to hire creative toolmakers that want to be the best in their craft. We believe this will enable us to achieve our mission of enabling the world to better solve it's problems. We are looking to expand our Engineering Leadership team.
As the Head of Platform Engineering, you'll lead the engineering group responsible for Notion's API and the broader platform surrounding it. Apply here: https://grnh.se/ef56d33e3us
As the Head of Core Product Engineering, you'll lead the engineering group responsible for the core of Notion's product—the docs and wiki experience, the editor, our mobile apps, collaboration tools, and more. Apply here: https://grnh.se/3a76669e3us
I was searching for research around the infection risk associated with various activities (outdoor vs indoor dining / gyms etc). Was hoping to see arguments motivated by contact-tracing data. Wasn't able to find much. Did anyone else have better luck?
You will not find much if any based on contact tracing data. Typical participation in CT isn’t remotely close to “I ate at x restaurant on y day”, and the entities tasked with CT are often inadequately equipped to handle the volumes/technical capacity of such a task.
There is, however, a wealth of literature using cell phone mobility, such as shown above.
The Wikipedia page describes what happens when someone does call. I wonder if that was a priori knowledge or if the writer called to see what would happen and then shared his findings :P
How do you build the prefixes when multiple words share the same prefix?
If my understanding is correct, the method is:
1. Find common search terms
2. For each search term (eg test, compute it's vector [1.23, 4.56...] and it's prefixes [t, te, tes...])
3. Store these in Qdrant as t->[1.23, 4.56...] , te->[1.23, 4.56...] , tes->[1.23, 4.56...] and so on. Here, each of the prefixes are used as point_ids
4. When a search query comes in, call /recommend and pass in the partial query as the point id