䷉Table: Companies in Knowledge Discovery (mostly Biology)

Send your thoughts via twitter or mail

A sporadically updated table of companies I come across. The main focus is on life sciences, knowledge discovery, domain specific search, automated reporting.

In random order.

Table published from Roam through ivywrite


kd/comment

JTBD

tellic

UIs not great, but good better granularity and molecular biology focused features and [[Named Entity Recognition | NER]]

I’m short on it.

Detect biomedical concepts in text data with PhD accuracy at Big Data scale

tellic-graph.png

support for ever-evolving vernacular, ontologies and identifiers

inform hypothesis generation and validation

Eagle Genomics

The platform applies knowledge graph, AI and unique valuation technology to reveal entities of interest beyond the tacit knowledge of a scientist or team of scientists to spark novel insights and drive innovation.

Cosmetics, Agritech, Food, Personal Care, Health Care

automated data scientist for biologists

Lens

Paper and Patent Search with relatively good Entity Resolution and Features

Golden

I do not understand Golden. Seems like a glorified wiki with extra UI sponsored by a16z VC $$$.

Nobody I know uses it. Possibly good for [[company search]]

Has some proprietory data for startups and funding rounds, so it's probably most valuable to VCs and Buzzfeed employees

privatized [[Wikipedia]] with good UI?

Elsevier's Services

I guess it has the problems of any [[designed by committee]] product offering

still have to research [[Elsevier]] offerings in detail

Reaxy

... enterprisey

Occamzrazor

Their development speed is questionable (started already in 2015 -- didn’t have product for first 3 years)

Most of their website is about their advisors and investors; pretty much nothing on the product

NLP discovery and drug target prediction for parkinson's

biomedical data omics

Linguamatics

**Apparently used by 19/20 Pharma Corps which makes it by far the most established product. **

Closed-off trials; sales teams; not for consumer

NLP for drug repurposing, [[Patent Search]], population health, quality measures etc.

biorelate

Seems to have prominent users

Sent a Galactic (their product) self-signup user demo to [[@Sean Glaze]] a while back - he said it was not very helpful...hard to tell

comprehensive and up-to-date curation of biomedical research

Lum.AI

Extremely stacked team, who authored many interesting papers I've read

Impoverished interfaces. Seems to not a self-serve solution, but mostly a consulting SaaS

Provide [[agriculture]] with insights into the changing research landscape and its impact on R&D priorities through NLP

focused on soil health, agriculture and food management

https://cellarity.com/

virtual cell simulation?

HealX

Drug discovery AI for rare diseases

https://www.springdiscovery.com/

longevity + ML

instabase

billion dollar company

document structuring + workflows; similar to hyperscience.com

https://www.expasy.org/

...

open source biomedical database

[[Wolfram Alpha]]

Wolfram is ramping on life science knowledge base

One of the most impressive systems I know.

Calculations on low level physical, mathematical phenomena.

Great for open question-answering (deployed by siri and google)

scientific computing, automated reporting, open-ended question-answering

[[Scite.ai]] source

Their whole "AI" should take six months, not three years.

It's marginally useful and [[> provenance tracking]] of [[citance]]s (citing a fragment of a paper) is interesting to us.

Good CEO, mediocre tech team.

[[Causaly]] source

Yiannis and Artur are a very strong CEO+CTO team

Seems they had some trouble to hire great senior talent for some reason (which is common).

The only life science knowledge discovery team I know that actually shows their UI (which is very well done)

Quertle

Has some interesting features, like bubble plot, but seems to be stagnant. The CEO doesn't seem technical and they don't have strong AI talent on their team

Pubmed Advanced Search

Actually somewhat powerful, but again only on the bottom-up semantics and not the top-down relation extraction level -probably because it's not deterministic and pubmed doesn't want to be responsible for noisy results

{{or: Semantic | Google}} Scholar

Google's UVP is "we have ALL papers". Semantic Scholar is pushing actually new features monthly now. Interesting, but only [[lowest common denominator apps]] that work for ALL scholars (horizontal AI tech; not optimized for specific workflows $$$).

Features are focused on horizontally applicable ML (like [[discourse tagging]] and [[Recommender System]])