CEO @ScribeHow. Here to help companies own their specialized intelligence & make their AI agents work. Alum @greylockvc @mckinsey @princeton
San Francisco, CA
Joined June 2010
- Tweets258
- Following346
- Followers1K
- Likes232
Jennifer Smith retweeted
😱 youtu.be/0kybbAvzBLo - a company spent $200M automating a process that cost $100M.
It is what Jennifer Smith @scribeceo, co-founder and CEO of @ScribeHow, says she heard from a healthcare company while advising enterprises on AI.
Before founding Scribe in 2019, before ChatGPT made AI mainstream, Jennifer was at @GreylockVC, where she interviewed more than 1200 executives about the problems keeping them up at night.
The pattern she kept hearing:
“I don’t know how my business actually works.”
Companies could see salaries, tools and outputs. But they could not see how work actually happened. That knowledge lived only in people’s heads.
That insight became Scribe: a $1.3B AI company with 6M+ users, used by 94% of the Fortune 500, and now past $100M in ARR.
In our new interview, Jennifer explains:
▪️Why most AI pilots fail before they scale
▪️Why more data can actually make AI agents worse
▪️How companies should turn internal know-how into an AI-ready asset
▪️ Whether AI will replace consultants, @McKinsey and knowledge workers
▪️The one human skill she believes will matter most
Her most contrarian point: enterprises do not have an AI model problem. They have a context problem.
“AI isn’t your solution until it understands your situation. That starts with your workflows.”
Full interview with Jennifer Smith from @HumanXCo in Amsterdam is now live on @aicryptominds YouTube➡️ youtu.be/0kybbAvzBLo
What is the biggest reason AI projects fail inside large companies: bad data, unclear ROI or lack of context?
"Knowing what to ask, how to find the information that matters for AI transformation, and THEN being able to build it successfully, is where all the alpha lies."
Agree with every word.
The next thing leaders need to realize is that they cannot and will not find that information manually.
Town's $108M Series B announcement and Ramp's Glass announcement are clear signs that AI agents have a problem leaders should be paying attention to.
Agents can't create automate work and deliver the value they're supposed to without figuring out how companies work... and most businesses have no record of how they work.
I've thought about this problem for the last decade, and I predict we're about to see a lot more companies try to build in this category.
I wrote about which approaches will and won't work.
Companies love to talk about processes as if there’s one right way and everyone follows it.
But the largest dataset on knowledge work in existence (Scribe’s) says the opposite: the most common version of a process only happens 7% of the time.
Run it 100 times and you’ll get around 50 different versions. Train an agent on the most common one, and it’ll match how your people actually work just 1 in 14 times.
That’s why we built Autocapture: so your people and your agents learn how work really gets done.
See it in action:
scribe.com/capture/autocaptu…
The most valuable thing at your company isn't in any system. It's the stuff your best people just know and have never had time to write down.
It’s your company’s specialized intelligence. And until today, there's never been a system built to help you own it.
We just launched Autocapture at Scribe: it turns how your teams actually work into documentation, automatically. No one has to press start or stop anymore.
What's never been encoded becomes visible. Structured, actionable guides, for people and agents.
See the announcement: scribe.com/library/autocaptu…
Every AI transformation plan has the same first move: connect the knowledge base, import the transcripts, let the model sort it out.
But more inputs isn't the same as better inputs. What matters is whether any of it reflects how your company actually works today — most of what gets written down is already stale by the time it's uploaded.
We've been building toward a different answer. October 6:
scribe.com/event-registratio…
Last week at @HumanXCo , someone asked: "If you started an AI company right now, what would you build?"
Everyone named uniquely human things: Enjoyment. Family. Learning. Rest.
One way to create that space: work differently. When AI takes the manual work off someone's plate, they get to focus on the things that energize them.
I believe human potential is the strongest force in the universe. Right now, we're squandering it, burying it under work that was never worth any person's time.
On October 6, we're announcing something from @ScribeHow built to change that.
scribe.com/event-registratio…
Every exec I talk to is saying the same thing: “AI knows everything except how my company works.”
That's because it runs on general intelligence that's rented – same labs, same models, but no unique edge. The companies that harness their specialized intelligence now will build a generational advantage that’ll only compound.
Join us on October 6 to learn how.
scribe.com/event-registratio…
Your company's most valuable asset isn't codified in any system.
It’s your specialized intelligence: the judgment calls, the workarounds, and the exceptions that people make every day without thinking twice.
When they leave the company, it leaves with them.
On Oct 6, we're changing that. Tune in live at 1pm ET / 10am PT.
getscribe.how/announcement20…
Jennifer Smith retweeted
Your company's most valuable asset isn't codified in any system.
It’s your specialized intelligence: the judgment calls, the workarounds, and the exceptions that people make every day without thinking twice.
When they leave the company, it leaves with them.
On Oct 6, we're changing that. Tune in live at 1pm ET / 10am PT.
getscribe.how/announcement20…
Agree with every word of this. Context creation isn't a nice-to-have on the side of transformation. It's the foundation everything else has to be built on for you to get value out of AI.
Companies need to harness their specialized intelligence — the proprietary context that makes them unique — and turn it into an asset they actually own, rather than something that lives in people's heads.
Databricks CEO @alighodsi went off on @a16z pod about enterprise AI adoption:
"They're just so far behind in the adoption curve of actually automating things and getting value out of this stuff."
Ali says most companies are still just using chatbots. There's hardly any agentic transformation.
Why is that?
"The models are smart enough, but they just don't have the context that exists inside of any organization."
"They have not been in every meeting. They don't know what's in everybody's heads. They don't know all the processes."
"If you just fused that and gave that context into the AI models...there's so much productivity gains you could get for any organization on the planet."
Context Creation is the biggest opportunity in AI right now.
Media is full of stories meant to make us mistrust and hate each other.
Here’s a story of a 9/11 hero - a 24yo kid who, trapped in the towers, selflessly decided to save almost a dozen people rather than himself. He found a path to safety and instead chose to go back in to rescue more survivors.
On that awful day - and all of the ordinary days - there are countless acts of human good.
Of people who put others before themselves.
Of the kindness of the human spirit.
Let’s drop the rage clickbait and share those.
Jennifer Smith retweeted
The jobs apocalypse is postponed. An AI jobs boom is here
According to @TheEconomist, AI is actually proving to be a net job creator in the US, easily generating over 1M new positions (from data center construction to AI engineering) to offset back-office layoffs.
While routine admin and customer service roles face real disruption, the overall labor market is proving resilient.
economist.com/finance-and-ec…
100% of our team is active on AI tools. 71% of our org is active in Claude alone daily.
Here's what got us there:
1. We set the expectation that AI fluency is a baseline expectation at Scribe, and a big part of how we move faster and win. Usage spiked every time we celebrated AI wins at All Hands. People saw what was possible, then went and did it themselves and got celebrated for it.
2. We didn't insist on building everything in-house. Buy was often the right call. No tooling to maintain, faster time to value.
3. We gave everyone access with limits by team. While a person knows their own workflow and where AI can help best, we didn't want unintentional token waste. We shared efficiency best practices across the org like delegating to sub agents and watching the session length to make this work.
Companies have years of data that exists nowhere else on the internet.
Thomson-1 is the first example. It's trained on decades of proprietary content, technology, and domain expertise no other company can match. The result is a model Thomson Reuters fully controls, without the heavy inference costs of typical frontier models.
The model has been trained on less than 10% of Thomson Reuters content so far, and is still on par with the latest frontier models across a range of tasks.
Now that the base open weights AI models are getting far better, and post training infra is becoming more mature and commercialized, there are going to be all new plays for companies that have large amounts of data to have their own models.
Licensing data for external model training was previously the only play if you had a large corpus of information, but you can now reasonably go and train your own models as well without incurring the cost and complexity of competing with the labs on research.
The general purpose frontier models will still have a leg up in broad areas due to the ability to handle the widest set of tasks, but you can absolutely see a future where we have far more models than today across every vertical and domain.
A few observations on AI transformation in the enterprise from recent conversations:
* “Agentic” does not necessarily mean fully autonomous. The more durable implementations are combining deterministic workflows with AI steps, especially where reliability and debugging matter.
* The real work is selecting the right workflows to automate, integrating fragmented data, setting governance, and helping teams change how they work.
* Security and privacy are hard requirements. Clear controls over approved apps, user privacy, and employee review materially affect adoption.
* Human-in-the-loop is still the default for consequential outputs. One team put it well: AI is great for the middle-to-middle. Brainstorming and final review should stay with a human.
* Companies are becoming more selective after broad experimentation. Teams are moving away from unlimited AI-tool purchasing (no more "let a thousand flowers bloom") toward measurable ROI and business case requirements.
* Token efficiency is becoming real operational work: model routing, shorter contexts, sub-agents, and centralized observability layers.
It's the same planning fallacy - we overestimate how much we can do in the short-term, and underestimate how much we can get done long-term.
It's so common in tech, it has a name - Amara's Law.
People see how sophisticated the models are and get happy eyes about what's possible....and then get hit by the reality of actually making that work in the enterprise. It requires so much more than just great models.
This chart from @McKinsey is very telling and reflective of human nature and organizations. Impact of AI on organizations is slower than anticipated, as it is with almost all technology. However, expectations continue to be aggressive.
We overestimate AI's impact in the short term and underestimate it in the long...
Had a great conversation with @DataFramed about why agents haven't dramatically changed knowledge work (yet) and why companies must harness their specialized intelligence first.
The models are PhD-level at this point, but none of that training tells them how your company uniquely works. That's what I mean by specialized intelligence - the accumulated decisions, judgments, and exceptions you've built up to serve customers and ship product your way. It's real IP, but it's still stuck in people's heads instead of being something that you can apply to your AI to build a generational advantage for your business.
At Scribe, we're working with 90K+ companies (including 50% of the F500) to help them harness their specialized intelligence and turn it into an asset that keeps compounding.
Listen → datacamp.com/podcast/what-do…
AI agents don't need to get smarter. They need to learn how your company works.
Jennifer Smith (CEO, Scribe) on DataFramed: why generic AI isn't a moat.
Listen → datacamp.com/podcast/what-do…
#DataFramed #AIatWork