2023-08-14 - KDB Project Call Meeting Minutes

alvin-c-shih commented 1 year ago

Date

Monday, August 14, 2023 - 9:30am ET / 2:30pm UK

// Second Monday of every month

Attendees

Fullname	Affiliation	GitHub Username
James McLeod	FINOS	@mcleo-d
Aaron Davies	Morgan Stanley	@adavies42
Peter Gyorok	KX Systems	@gyorokpeter
Alvin Shih	Morgan Stanley	@alvin-c-shih
Jo Shinonome	ExeQution Analytics	@jshinonome
Ross Duffy	Morgan Stanley	@rdffy
Gary Davies	Data Intellect	@drgdavies
Connor Gervin	KX Systems	@Connor-Gervin
Conor Twomey	KX Systems

Meeting notices

FINOS Project leads are responsible for observing the FINOS guidelines for running project meetings. Project maintainers can find additional resources in the FINOS Maintainers Cheatsheet.
All participants in FINOS project meetings are subject to the LF Antitrust Policy, the FINOS Community Code of Conduct and all other FINOS policies.
FINOS meetings involve participation by industry competitors, and it is the intention of FINOS and the Linux Foundation to conduct all of its activities in accordance with applicable antitrust and competition laws. It is therefore extremely important that attendees adhere to meeting agendas, and be aware of, and not participate in, any activities that are prohibited under applicable US state, federal or foreign antitrust and competition laws. Please contact legal@finos.org with any questions.
FINOS project meetings may be recorded for use solely by the FINOS team for administration purposes. In very limited instances, and with explicit approval, recordings may be made more widely available.

Agenda

(5 min) Convene, roll call, welcome new people.
[ ] Display FINOS Antitrust Policy summary slide
[ ] Review Meeting Notices (see above)
(5 min) Approve previous meeting minutes.
- Approved by Acclamation
(5 min) KX CLA Updates
- Reports that the CLA not getting approved due to lawyers getting impatient with poor UX.
- Tickets rather than a portal tailored specifically for this.
- James will look into getting this particular one through.
(20 min) kuki, the K Ultimate pacKage Installer (Jo Shinonome)
- https://github.com/jshinonome/kuki/wiki/getting_started
- can install packages straight off GitHub
- somewhat reminiscent of Emacs straight.el
  - https://github.com/radian-software/straight.el
- utility to initialize a project with correct structure:kuki --init
  - possibly analogous to nbdev : https://nbdev.fast.ai/
- supports for building plugins
- Jo also hosts a registry!
  - https://kuki.ninja/
- also includes process launcher and test runner!
- question about whether enterprises with scanning tools for PyPI and NPM would integrate kuki as well
- intention is that it could be internally hosted within a bank!
- good for banks that don't already have a package installer ecosystem
- useful for sharing small parts of larger code bases like TorQ
- next steps?
- full ticket architecture examples
- example plugin - possibly regex
- ideas for low-level functions to wrap
- _exit(2), usleep(3), setrlimit(2)
- other plugin ideas
- database drivers, character set conversion
- question about multi-repo support (a la Anaconda)
(20 min) LLMs and kdb
- demonstration of http://kdb.ai
- team of 30 people working on it
- going up against the likes of Weaviate, Pinecone, Chroma
- time will tell if Microsoft finds a way to dominate this space through Office 365 Copilot, etc.
- separate product from kdb
- transitioning from a services company to a product company
- additional operators beyond mmu for nearest neighbors, etc.
- 8 enterprise clients registered for private preview
- Private Preview 2 dropping on September 12th
- brings together structured data plus unstructured data
- use case: event processing based on news headlines
- applicable to: travel, finance, etc.
- benefit of explainability / citations
- pricing still being worked out
- should be easier to integrate with existing kdb plants
- customizing kdb.ai is mostly done in Python for now
- will be interesting to see if results can be tuned based on date range / diversity
- de-facto standard benchmarking suite yet to emerge
- seeing if there's demand for multimodal vector databases
- kdb pedigree might not be a huge door-opener
- time series people are often in a different silo from the full text people
(5 min) AOB, Q&A, Adjourn.