Sierra Michelle Dowling Poudel

PR at @glean. Thoughts are my own.

[Molly O’Shea @MollySOShea]

Arvind Jain (@jainarvind), CEO of $7.2B @glean, aka the knowledge layer for enterprise AI that recently crossed $300M ARR.. says:

"Open source models are gonna dominate AI inferencing."

"You'll see in the next two years we'll shift to where it's almost no open source to where it's gonna be almost all open source."

The 2 biggest needs for enterprise AI:

[Sumanth @Sumanth_077]

Small Language Models are the Future of Agentic AI!

Glean just released Waldo - a 30B agentic search model that runs before frontier LLMs.

Search is where most agentic work begins. You ask about a project, customer, process, or decision. The agent searches internal docs, reads results, refines queries, searches again. Sometimes one search. Often several iterative loops.

Get the search wrong - miss a critical document, surface irrelevant results - and the entire response fails. Search planning is the foundation for AI Agents.

But frontier models are doing two different jobs at once. Search planning (which queries, when to stop, is there enough evidence) and synthesis (reasoning over results to generate an answer). The first job is pattern matching. The second needs deep reasoning.

Waldo splits these. It's a 30B MoE model built on Nvidia Nemotron 3 Nano that handles just the search planning layer. It runs first, decides which queries to run across Glean Search, Employee Search, and Web Search, determines when it has enough context, then hands off to the frontier model with retrieved context already in place.

Key architecture:

• Run Waldo first, before the frontier model. The alternative (sub-agent design) would require the frontier model to call Waldo as a tool, wait for results, then respond - two frontier calls. Running Waldo first cuts it to one.

• Training Phase 1 (DPO): The model learned when to search, when to stop, and when to hand off from production tool-use patterns. The training data captured which tools were called, in what sequence, and whether the plan succeeded.

• Training Phase 2 (RL): The model was trained against production queries and rewarded based on document recall - whether its searches surfaced the same documents that appeared in successful final answers. This refined its ability to find relevant documents in fewer search iterations.

• Results: 10x faster per call (250ms vs 3s). Half of queries run on this fast path.

The pattern: specialized small models for focused, repetitive tasks. Frontier models for reasoning and synthesis.

Waldo proves small language models are faster, cheaper, and just as effective for repetitive, focused tasks.

I've shared the link in the replies!

[Sierra Michelle Dowling Poudel @sierradowling]

Check out @JayaGup10 and @jainarvind discuss the reality of enterprise AI: context > models

https://t.co/DmEW6MGJSd

[Jaya Gupta @JayaGup10]

Really awesome getting to chat with @jainarvind !!!

[Tony Gentilcore @tonygentilcore]

Building an enterprise brain is harder than most people realize. It's not just ingestion, it's figuring out what entities matter, how they relate, what to remember, and what to forget. We've been deep in this at @glean.

More thoughts threaded below.

[Glean @glean]

We’re diving into what it really takes to make AI work in the workplace. If that sounds like sessions you want to see at @SXSW 2026 – vote to get them on the agenda!

🔷Enterprise AI at scale with @jainarvind, CEO at Glean, David Beitel, CTO at @Zillow & T. Alexander Puutio at @ForbesTech — https://t.co/HSjGaYbBPJ 
🔷Getting the most out of your enterprise data with AI with @jainarvind, CEO at Glean, & @barisg, VP of AI at @Snowflake — https://t.co/TBLWfhAEDa 
🔷AI Alignment and distributed work with @jainarvind, CEO at Glean, & Dan Spaulding, CPO at @Zillow — https://t.co/vG3jcf0lT1 

Your voice is important! 🗣️ Community voting makes up 30% of SXSW final selection. Voting closes August 24, so don’t miss your chance!

[Arvind Jain @jainarvind]

AI models today are incredibly powerful, but we’re only scratching the surface of what they can do to drive true business transformation.

Elon Musk's recent comments shed light on an important challenge: the limits of training data. But the real conversation should focus on the untapped potential in applying these models.

The technology is here, but the gap between capability and meaningful application is immense. That’s where the opportunity lies.

At @glean, we see this every day. The early innings of AI are already reshaping how people and businesses work. The best is yet to come.

Curious to hear what others are thinking about this…

[Arvind Jain @jainarvind]

“Without [trust], the promise of AI will stall, leaving businesses behind in the ‘Intelligent Age’” - @BetterUp's @Arobichaux via @wef

I couldn’t have said it better myself. Trust isn’t just a technological challenge - it’s a human one.

🧵 1/3
https://t.co/qQ0FALXcKu

[Glean @glean]

"Glean is the fastest way for you to bring useful enterprise AI to all of your employees. We make it easy for you to make all these amazing AI models work on your enterprise data in a safe and secure way."
- our Founder & CEO, @jainarvind with @AnsafK on the Escape Velocity podcast

Give it a listen 👇

[NYSE 🏛 @NYSE]

The NYSE welcomes @PureStorage in celebration of its 15th anniversary of founding! $PSTG

[Jon Fortt @jonfortt]

I speak with @PureStorage CEO @Pure_Giancarlo about $PSTG fiscal second quarter earnings results and guidance, enterprise demand, and the cost of storage. https://t.co/8h1nHlZpyV

[Miroslav Klivansky @klivansky]

Q2: There's value for #enterprises who make the investment to build their #AI expertise. Lots of experimentation and failing fast is a good thing. We suggest building out an AI Center of Excellence as the best place to learn #eweekchat

[Everpure @EverpureData]

Happening now! Join Pure's Miroslav Klivansky for a conversation about how the enterprise is grappling with the rise of #AI.

@klivansky

https://t.co/QDbYaJRvgs… #eweekchat

#eweekchat

[Miroslav Klivansky @klivansky]

A1 for #eweekchat: In short, yes! We're still not sure exactly what will stick and where #GenAI will have the most impact, but that it'll have impact seems close to certain

[Sierra Michelle Dowling Poudel @sierradowling]

Check out @codyhosterman on @theCUBE now to learn more about #CloudBlockStore for Azure VMware Solution at #VMwareExplore ! @PureStorage https://t.co/t4xceN3veL

[Everpure @EverpureData]

Don't miss the chance to listen to our Sr. Director of Product Management - FA, @codyhosterman, who will be on @theCUBE this afternoon (4:30pm PST) at #VMwareExplore.

Live stream link: https://t.co/XrJUDbGI6b

#data #DataStorage #AI #ML #DataAnalytics

[Sierra Michelle Dowling Poudel @sierradowling]

Krithika Bhat, CIO at @PureStorage, is speaking about her experience as a female immigrant in technology. Listen now! #womenintech #womenwhocode #womeninstem #WTGC2023 @womentechnet https://t.co/S4YA3WvNaG

[Tech Field Day | #AIIFD5 June 10 - 11, 2026 @TechFieldDay]

Live now, it's @PureStorage presenting at Storage Field Day 24! #SFD24 #Pure1

https://t.co/As3BC812YR

[Reuters MOMENTUM @ReutersMomentum]

"The total data storage projected in 2025 is set to exceed 200 zettabytes. That’s 200 billion terabytes."

Ajay Singh, Chief Product Officer at @PureStorage Storage speaking on managing data growth at #ReutersMomentum.

Tune in now! https://t.co/wXrSplXvnd

#ReutersMOMENTUM

[PrimeImage Media @PrimeImageVideo]

Live from VMware Explore! #TFDx