pygraphistry

ntil convergence has already outperformed human review for us . (Doesn't replace, but does find things we miss, and catches many we do see earlier).

r the louie.ai umbrella, piloting genAI extensions

heapo GPUs just by being data parallel with high saturation per step: https://github.com/graphistry/pygraphistry/blob/master/graph... . Despite ping-ponging a ton because cudf doesn't (yet) coalesce GPU kernel calls, V1 competes surprisingly high, and is easy to maintain & extend.

u -> gpu. It already has sentence_transformers built in. Due to our work with louie.ai <> various vector DBs, we're looking at revisiting how to make it even easier to plug in outside embeddings. Would be curious if any patterns would be useful there. Prior to this thread, we weren't even thinking folks would want images built-in as we find that so context-dependent...

epresentation, and we are now working on adding the first extensions to support on-the-fly switching to more graph-native indexing as needed. Ex: https://github.com/graphistry/pygraphistry/blob/master/demos...

drilling down, navigating, reclustering, etc. - A useful 'aha' is thinking of modern information visualization as trying to optimize some sort of time-to-insight through a sequence of visual interactions. So each view must be information dense for visually revealing certain insights, and make it easy to get to the next set of visual Q&A. - Ex: When the entities are the interesting thing wrt questions, being able to drill down into individual nodes/edges into great dedicated views becomes important, so graphs should often be multimodal. And if the relationship aspect is unimportant... then using graph view hurts more than it helps. - From optimization perspective, it now makes sense to specialize for specific domains. Maybe what is needed is more of a small diagram, and not actually investigating a lot of relationships. Or a graph of subway stops, which has additional visual considerations. For a website, a sitemap navigation vs clickstream product analytics view would likewise need A good analogy is a map. Sometimes exploring Google Maps is great, and you drill into a business inspector sidebar or down to a street view. But other times, it's better to have the map embedded into Yelp.com restaurant entry when you just need a quick view of mapping information as part of some broader context. Or you don't care about that map at all and can skip it. Given all that.. it's interesting to revisit asking... what is the 'content' job to be solved? What kinds of content lean towards graph, and which don't?

orpheus & graph AI sides for end-to-end GPU pipelines with our bigger customers (cyber, ...). More recently, to make this kind of tech easier for analysts who are traditionally stuck with Splunk/Kibana/etc style UIs for investigations, we have been launching louie.ai (genAI-first notebooks) with various customers Hopefully now it makes sense why we don't go far with candidates who can't have conversations on these things, such as how they are built, how they get used, and where they are going . It is ok-but-not-great on conv 1, but weird by conv 2. And as a CEO, far from what I look for in someone in a senior/leader role who is supposed to be looking ahead.

s to entities, like "(signup)-> (user agent)" + "(signup)->(ip)" + ... . But other times its annoying, so they rather just see "(ip:node)<-[signup:edge]-->(useragent:node)". Interestingly, for really wide data, having an explicit "hypernode" (event node) does increase the number of nodes... but significantly decreases the number of edges b/c avoids drawing really big connected components. Formally, add |events| nodes, vs multiply edges by |number of columns|^2 . If you want to play with it, try "graphistry.hypergraph(pd.read_csv('http://blah.csv'))['graph'].plot()" in http://github.com/graphistry/pygraphistry (or just to get the ._nodes and ._edges dataframes out) Nowdays, we also do a lot with embedding arbitrary data and exploring similarity graphs that way (use k-nn for inferring similarity edges), so not just across entity links but also say timestamps, $'s, and byte counts. Any dataset embedding can thus be thought of as a projection of the hypergraph, and supporting more interesting columns than just entity columns (categoricals). So in the above example, `graphistry.nodes(pd.read_csv('http://blah.csv')).umap().plot()` and we'll infer the `._edges`. All the nodes in an embedding view are essentially hypernodes, and instead of connecting them to entity/attribute nodes... project those out, and just connect the hypernodes to one another. This gives a very different way to think about stuff like explainability of AI :)

not discussed like time), GPUs, clusters, data pipeline integrations, etc. Imagine having 100K cyber alerts aggreated every day or a 50M users/transactions and wanting to score them... It's nowhere as easy (yet) as something like adding a layer to BERT like NLP people can do. If that's more your speed: - We're working on OSS automl for graph ai, trying to get typical internal team pipelines to go from events data to decisions & UIs in a few lines. First out was for UMAP and we've been pushing on GNNs more recently, http://github.com/graphistry/pygraphistry . Elsewhere, we're also working on the MLOps side and some key modeling scenarios. - ... both graphistry + almost all our customers & partners are hiring here! If you're into data (analytics/mlops/dataeng/ds), or general js/python fullstack dev, a lot happening here for missions like supply chain, cyber, fraud, & misinfo. Would love to chat! EDIT: Another two things I've found interesting: -- GNNs are probably most researched by folks in material sciences (chem, physics, ...), and it's really changing things like protein folding, and maybe next biggest on social networks. We see equal practicality for other key problem (cyber, fraud, supply chain, ...), but we're seeing much less academic work in these, and I think that's b/c academics are at a significant data disadvantage vs most industry teams. So even though big interest outside of the academic team areas, it's quite early days. -- Industrially, we're seeing GNNs promoted primarily by graph databases.. but almost all graph databases are CPU-based vs GPU-based, and in our polls of commercial GNN usage, typically not as part of a graph DB pipeline but something like regular data lakes (parquet, ...) feeding into regular scalable GPU compute tiers