The AI Performance Company, on a mission to make AI better for everyone and build ML technology to drive responsible business results. Tweets written by humans.
We’re thrilled and honored to say we’ve been selected by
@CBinsights
as one of 2020’s top 100 AI startups! Huge thanks to our whole team for getting us there! 👏🏻👏🏻👏🏻
Announcing: Arthur Shield. 🛡️
The newest in our suite of AI monitoring tools enables companies to detect and block critical risks that come with deployed LLMs, including sensitive data leakage, hallucinations, and toxic language generation.
Learn more:
We’re thrilled to announce that we’ve raised $42M in Series B funding!
Led by
@AcrewCapital
and
@greycroftvc
, this round will allow us to accelerate our mission to make AI work for everyone.
• Blog post:
•
@TechCrunch
story:
Introducing Arthur Bench! 📊
Bench provides teams with a single touchpoint for all of their LLM performance evaluation, helping them compare different options quickly, easily, and consistently to determine the best fit for their application.
Learn more:
GPT-3 is a transformational technology that businesses will be able to buy from OpenAI. Arthur’s Chief Scientist
@johnpdickerson
sat down with
@business
to talk about the model’s core bias that may leave businesses exposed to legal and reputational risk.
Recently, ML Engineer
@maxcembalest
created an “Ask Arthur” chatbot using
@LangChainAI
,
@Gradio
, and native Python packages. 🤖💬
Get a walkthrough of how he constructed it and even give it a whirl:
Great read on using responsible AI to build better businesses:
Doing good + doing well = what
@lo_toney
calls "front-end ethics" -- what we're all about at Arthur.
More on how to mitigate algorithmic bias here:
TOMORROW: We’ll discuss all things
#LLMs
with an all-star lineup at our next Ground Truth event!
@alexrkonrad
, Senior Editor at
@Forbes
, will moderate a panel including:
🔹 Angela McNeal - Ex-Palantir AI, Co-Founder & CEO
🔹
@bobvanluijt
- Co-Founder & CEO,
@weaviate_io
🔹
Arthur is donating $2,500 to Breathe India, a campaign raising money to address the devastating health crisis in India caused by COVID-19. We are matching donations up to an additional $2,500. See thread for details.
Last week was one for the books!
💸 Announced Series B raise
💻 Launched redesigned website
🏙 Moved into new HQ
🎉 Got the whole team together for Arthurfest
What’s next? Check out what our CEO
@apwenchel
has to say about Arthur’s next chapter:
☁️ We’re excited to bring our class-leading model monitoring solution to even more teams through our strategic partnership with
@awscloud
, which will allow AWS customers to purchase and deploy Arthur directly from the
@awsmarketplace
.
Learn more:
Excited to host
@polynoamial
, Research Scientist at
@MetaAI
and creator of the first
#AI
bots to beat top humans at two-player no-limit and multiplayer poker, for the Karthik Rao Anniversary Lecture at today’s
@itsArthurAI
#ML
offsite!
Our ML Research Team hosts a discussion group every Friday about the latest and greatest research papers on
#XAI
, ML fairness, and
#ResponsibleAI
. We decided to do it on Clubhouse this week, and we'd love it if you joined us tomorrow:
Exciting news from the team here at Arthur! We’ve been selected to join the 2020
@FinTechLab
lineup to help solve real problems and drive innovation for the finance industry.
We're excited to announce that Arthur has been named a 2021 Gartner 'Cool Vendor' in AI Governance and Responsible AI! 🎉
Read more about the announcement on our blog:
When the world suddenly changes, your models suffer. Read about how Arthur helps leading companies like
@Humana
navigate the challenge of AI during COVID19 in today’s
@WSJ
:
Thrilled to be hosting “The Future of LLMs” next Thursday with an absolutely stacked lineup! 🎉
Don’t miss hearing from leaders at
@LangChainAI
,
@MosaicML
,
@weaviate_io
, and more about building successful LLM companies.
🔗 Registration link:
It was a full house last week at Arthur HQ for “The Future of LLMs”! The energy in the room was palpable, and it’s exciting to see how much the AI community in NYC is growing.
Huge shoutouts to our talented speakers:
@hwchase17
of
@LangChainAI
; Angela McNeal, formerly of
We sat down with our latest research fellow Sahil Verma to talk about his exciting work at Arthur - read on to learn more about counterfactuals and Explainable AI!
We are so excited that our co-founder
@lizjosullivan
has been named one of the
#100BrilliantWomen
in AI Ethics this year! We couldn't be more proud of her work combatting algorithmic bias and unethical uses of AI.
Scenes from our all-team offsite,
#CampArthur
💜 in the Hudson Valley this week. There's nothing like spending a few days in the mountains together building connections, sharing meaningful conversation, and dreaming up creative ways to solve big problems. Plus, all the s'mores...
Want to get a behind-the-scenes look at the life of an Arthur engineer? We're launching the Arthur Engineering Blog to share some of the interesting problems we're solving, from picking a
#k8s
orchestrator to building a delightful Python SDK. Take a look:
Our CEO,
@apwenchel
, will be speaking with some Responsible AI all-stars after a screening of
@CodedBias
at
@UofMaryland
later today—tune in to learn more about the consequences of algorithmic bias:
Today, CB Insights released their “AI 100” ranking, showcasing the 100 most promising and innovative private AI companies from around the world—and we are thrilled to be included in the list for the second year in a row.
Thrilled to be launching “Ground Truth,” our new series featuring conversations with the brightest minds in
#AI
and
#ML
! 🧠💡
First up on March 23 is
@radcummings
, who is known for her work in data privacy, ML, and public policy.
Learn more and register:
Our co-founder
@lizjosullivan
weighs in on the risks of releasing GPT-3,
@OpenAI
's latest natural language generation model. GPT-3 is riddled with bias—Liz and other experts argue this must be addressed before it is made commercially available:
Who better to spotlight this week than
@shaneapollo
, who just celebrated his first Arthurversary?
Shane’s not only hilarious and super smart, but he’s also a fashion icon whose style can be best described with a phrase he coined himself: “humble swagger.” 😎 You rock, Shane!
“We regularly speak with ML executives from Fortune 1000 companies and one of their biggest concerns... is model behavior in production. The Arthur platform is by far the best solution we’ve seen for AI monitoring and transparency.”
—
@Work_Bench
VC
Team offsites this week: sales and marketing headed to Chicago and the engineers hosted a hackathon in NYC! So much fun meeting and learning from new coworkers in person. 💜
Congratulations to
@Sahil1V
,
@keeghin
, and
@johnpdickerson
on their award-winning survey paper on counterfactual explanations! We recommend reading the full paper, but you can also find a summary of its key points on our blog:
NYC ✈️ NOLA
We're thrilled to be attending
#NeurIPS2022
in just two weeks from today!
Head to to learn more about our ML R&D team, our recent research and publications, and some fun after-hours events we're hosting as well.
Hope to see you there!
Such an uplifting and inspiring day today at
#womenterprise23
, hosted by our friends at
@Work_Bench
.
We’re so lucky to be part of such an amazing community of women in enterprise tech—and we already can’t wait for next year! 💕🙌
If you’re in
@Montreal
this weekend, check out our Chief Scientist
@johnpdickerson
’s talk “Robustness, Privacy, Fairness, & Credibility?Pushing the Boundaries of Economic Design with Deep Learning”
@CIREQMTL
.
Issues of fairness & privacy arise in complex systems where... (1/2)
Fortune 500 companies know that the need for model monitoring is pressing, and can't be farmed out to a company whose focus is elsewhere. Thanks
@jerseejess
for the shoutout in
@TechCrunch
today!
Large language models come with large risks—and companies must be able to detect and protect against them.
🎥 Watch our CEO
@apwenchel
’s full appearance on
@BloombergTV
:
🛡️ Learn more about Arthur Shield, the firewall for LLMs:
Amazing
#TeamArthur
offsite this week in Boulder, CO! It was great to unplug for a few days of team building and knowledge sharing (plus some arcade games and delicious BBQ, of course). Feeling motivated and excited for all that’s ahead in 2022! 💜👏🏔
Want to learn how to set up your
#ML
pipeline for post-deployment success and help your business achieve operational
#AI
maturity?
Our very own
@hellovictoriav
is speaking today at the
@lesbiantech
#PrideSummit
! Learn more and register here:
The vast majority of algorithmic decisioning systems in use today were not built with consideration for bias or explainability. This needs to change.
@itsArthurAI
provides the tools data science teams need to AI responsibly.
#ExplainableAI
#bias
#ml
#ai
Arthur has been named one of the
@BuiltInNewYork
2022 Best Places to Work! 🎉🏆🙌
We're honored to be recognized & celebrate what makes us truly great: our team.
PS: We’re hiring! Join Team Arthur 💜 Open roles:
#2022BuiltInBest
"I really want to work at a startup, but..."
Our CEO Adam Wenchel shares the reservations we often hear at the end of this sentence -- and the two commitments we've made at Arthur that challenge "typical" perceptions of startup culture:
Fun
#FAccT
: This is the second year in a row that an Arthur Research Fellow has won Best Paper at the
@FAccTConference
!
Last year,
@ziebrah
won for "Epistemic Values in Feature Importance Methods: Lessons from Feminist Epistemology." Give it a read here:
Psyched to be hosting our second
@AITinkerers
NYC event later this month! 🥳
Sign up here for an evening of awesome demos and networking with other active AI/ML builders:
Had so much fun up at Bear Mountain for the
@Work_Bench
#LFGNYC2022
retreat earlier this week! Amazing views, great lessons, and even better people. 🏔🔥👏
We were impressed by all of the fantastic research on
#ResponsibleML
and algorithmic fairness at
@NeurIPSConf
back in December—we summarized some of our favorite papers, workshops, and keynotes in our latest blog post:
Today and every day, we’re proud to do our part to
#BreakTheBias
in AI and beyond. Happy International Women’s Day to the women of Arthur—and our community of customers and partners—who play such an integral role in helping us do so! 💜
#IWD2022
Follow along in this thread for highlights from today’s Bringing AI Model Monitoring to a Healthcare Setting event at the
@FAccTConference
.
#FAccT21
👇
"When algorithmic interventions in health care fail, they often do so disproportionately, leaving the most at-risk communities to suffer." Our co-founder
@lizjosullivan
and Parity AI CEO
@ruchowdh
wrote about the risks of applying unfair models to the Covid vaccine rollout.
"When it comes to decisions that amount to life or death, opaque technology can enable them but should never decide," say
@ruchowdh
and me in our latest op-ed on algorithms and the *many* ways they can fail us. Thank you
@ozm
for the opportunity!
The Arthur team spent last week in Vancouver at
#NeurIPS2019
! In case you couldn't make it, our Chief Scientist
@johnpdickerson
wrote down a few thoughts on his experience at this year's conference, and its trends and trajectory for 2020 and beyond...
Want to meet Arthur!? Schedule a demo with one of our experts to see the AI monitoring, explainability, and bias analytics platform in action: .
Check out why ML Practitioners at companies like
@Humana
trust Arthur to ensure responsible AI outcomes.
The one and only Madeleine is our
#TeamArthur
Spotlight this week! From keeping everything running smoothly in the office to planning the most incredible team offsites, she's on top of it all and we don't know what we'd do without her.
Join the team:
Together, Arthur and
@algorithmia
provide a powerful set of tools that give you complete control over your production AI. Today, we announced a partnership that makes integrating Algorithmia and Arthur easier than ever:
AI models can’t use context or common sense to infer why behavior is different when it suddenly changes. Do your systems have all the right monitoring in place so you can know the moment they fail?
Proud to be selected by the Joint Artificial Intelligence Center to provide critical AI performance solutions for the Department of Defense. Arthur's exclusive capabilities in CV, NLP, & bias mitigation will support AI for DoD operations and missions.
It’s a MAD world & we love every minute of it. Find us in the 2021 Machine Learning, Artificial Intelligence and Data (MAD) Landscape by
@mattturck
@john_d_wu
Find us: ML & AI ➡️ Deployment & Production. A small box with a huge impact on business success
What happens when
#AI
meets the real world? It doesn’t always behave as it was trained to.
That’s why we created Arthur—to help enterprise teams monitor, measure, and improve
#ML
models in production and achieve better results across accuracy, explainability, and fairness.
Thrilled to officially welcome two new leaders to the Arthur team! Tom & Hillary's work will be instrumental in helping us deliver on our mission to make AI work for everyone.
Want to join us? Check out our open roles:
AI is bringing big change to every industry, some good, some risky. Arthur CEO
@apwenchel
sat down with
@WSJ
to share his predictions about what we‘ll see over the next few years and how enterprises can prepare for it... read on for more!
On International Women's Day, we are celebrating the women in our lives and reaffirming our commitment to fighting gender bias everywhere. Thank you to the women of Arthur, the women of the AI community, and the women of the world.
#ChooseToChallenge
We're excited to join AI Governance International,
@AI_Guild
, and an incredible lineup of experts next week at the Responsible AI for Data Science summit. Tune in to hear the latest on AI observability and
#MLOps
.
The Arthur team will be at
#NeurIPS2022
in just one week, so we wanted to spotlight some of the research and events we’ll be involved with while we’re there! 👇
Our first Ground Truth event of 2024 is shaping up to be a great one! 🎉
Next Thursday at Arthur HQ, join us for “AI Crashes the Science Party” with speakers from
@AIatMeta
,
@nomic_ai
,
@genentech
,
@ai_seek
,
@AbridgeHQ
, and more. 🧪⚡️
Learn more:
Last week, our
#ML
team returned from NOLA full of knowledge, new connections, and beignets.
Now, they’re sharing their takeaways (and hot takes 🔥) from
#NeurIPS2022
on what’s right, what’s wrong, what’s hot, and what’s not in the
#machinelearning
world:
This story has everything! Bread photos, smart business, a fellow Arthur namesake, and more evidence that early detection and data validation MATTER when external conditions suddenly (or gradually) change.
New from
@MRKR
: In early March, King Arthur Flour noticed it had 600% increase in grocery-store sales almost overnight. They thought it was a blip.
#longreads
We had an absolute blast at
#NeurIPS2022
last week in New Orleans!
It was an inspiring week where we got to present our work, meet tons of great people, and even host a few events with our friends at
@IndexVentures
and
@abacusai
.
See you next year! ✈️
Heading into the holiday weekend feeling so thankful for this incredible Arthur team! So lucky to work with such genuinely lovely humans everyday, solving really interesting and meaningful problems together.
Wishing you and yours a restful, (ful)filling Thanksgiving! 🙏🍗🥧💜
BIAS IN ML SYSTEMS ARISES FROM A COMBINATION OF MANY FACTORS. WE HAVE TO STOP PRETENDING THAT "FIXING THE DATA" WILL SOLVE EVERYTHING. THERE'S NO SIMPLE FIX FOR THIS AND YOU SHOULD LOOK SKEPTICALLY AT ANYONE WHO CLAIMS THERE IS. BRING STAKEHOLDERS TOGETHER AND CONSIDER THE HARMS.
On Tuesday, members of the Arthur team will be presenting at
@AIESConf
!
Learn more about their paper, which explores the intersection and relationship between U.S. fair lending regulation and algorithmic fairness research:
We're ready for rooftop drinks! Counting down to our Enterprise AI Happy Hour, hosted with
@Work_Bench
-- Thursday at 5:30pm.
Join us for live music, craft cocktails, great food, and the best in the AI biz.
💜🍹🍕🎸🌟
Mark your calendars for June 9th! 📅
Our VP of ML,
@keeghin
, will be hosting a webinar about using ML to drive meaningful business results across financial services use cases like fair lending, fraud, and customer experience.
Save your spot here:
Product update: model versioning is live! Teams will now be able to:
🟣 keep track of models in validation or post-production
🟣 optimize ML models by easily viewing and comparing across versions
🟣 improve AI/ML governance
Machine learning experts are trying to improve fairness in AI, but is that possible without a shared language between engineers, laypeople, and policy makers? Our Chief Scientist,
@johnpdickerson
weighs in:
It's Friday, and you know what that means:
#TeamArthur
Spotlight! In addition to being a skilled engineer, Ike is also a proud alum of
@CMUEngineering
and a native New Yorker who you can probably find running through Central Park.
Join the team:
New feature alert! 🔐
Arthur is the first
#MLOps
observability platform to launch fully customizable Role-Based Access Control (RBAC). Learn more about how this unique capability will help enterprises strengthen data privacy and reduce compliance risks.
Our work with
@Humana
helps ensure their models are stable and explainable, providing insight for AI teams on where to focus efforts in a rapidly-changing world.
Scenes from
@satml_conf
! 📸
Teresa (
@tdatta_
) and Daniel (
@NissaniData
) are presenting their paper, “Tensions Between the Proxies of Human Values in AI,” tomorrow at 2:15. If you’re here, be sure to check it out!
#SaTML
We're officially a proud sponsor of The
#AISummit
New York, December 8th - 9th at the Javits Center!
We'll be on stage for three exciting sessions, and Team Arthur will be on site talking AI Performance and hosting demos in our expo booth.
DM for virtual passes! 📆🎟️🎉
“While there are... benefits to our accelerated use of AI during the pandemic, we’ve jumped in without the necessary checks in place to secure our safety & commitment to principles of justice — including avoiding... bias and preventing discrimination”
Arthur's in Austin! We'll be at the
@DataCouncilAI
Conference this Wednesday and Thursday, ready to talk all things MLOps and data science.
Learn more and schedule a meeting with our team here:
Podcast alert! 🎙 Check out the
@VentureFizz
Podcast featuring our Chief Scientist,
@johnpdickerson
:
He and
@kcline6
discuss innovation in academia, John's career and research, the path that led him to Arthur, and our mission to make
#AI
more transparent.
Big shoutout to
@ziebrah
(former Arthur Research Fellow!),
@geomblog
,
@scheidegger
, and Sorelle A. Friedler for their initial research on the topic, found here:
📆 Tomorrow, Arthur co-founders
@apwenchel
and
@johnpdickerson
will be featured in the AWS Startup Showcase where they’ll share their thoughts on generative AI, LLMs, and fairness in machine learning.
Register here:
Our co-founder
@lizjosullivan
spoke with
@alfredwkng
about the field of algorithmic audits and the work we're doing at Arthur to make AI fairer and safer.
“Once you’ve put it into the real world, a million things can go wrong, even with the best intentions,” O’Sullivan said. “The framework we would love to get adopted is there’s no such thing as good enough. There are always ways to make things fairer.”
Wise words from
@johnpdickerson
on the
@VentureFizz
podcast this week.
Hear more of his advice for aspiring entrepreneurs, as well as his thoughts on the pulse of innovation in academia and the career path that led him to founding Arthur with
@apwenchel
: