
What happened
Koch Disruptive Technologies, the venture arm of Koch, Inc., is making a strategic investment in Emanate, the San Francisco startup applying AI to industrial sales. Neither side would say how much.
Why it matters
Koch, which reports more than $125 billion in annual revenue and about 120,000 employees in over 50 countries, runs the refineries, factories and supply chains Emanate's software targets, so the deal pairs the startup with one of its potential users at scale.
WHO IT HITSIndustrial suppliers and their sales teams — the reps and account managers whose knowledge Emanate's software is designed to capture — face a tool pitched at their core job. Startups selling AI into heavy industry may also read Koch's backing as a signal that large industrial buyers are willing to test such software.
Summaries like this, in your inbox every morning.
Emanate first surfaced publicly in February, when Fortune reported it had come out of stealth with backing from a16z, Peter Thiel and Reddit cofounder Alexis Ohanian. Founder and CEO Kiara Nirghin is a Thiel fellow, a former Google Grand Prize winner and the youngest board member of the Google Impact Fund. The Koch investment adds a strategic backer whose own operations — refineries, factories, farm-to-market supply chains, software and logistics — look like the customers Emanate is courting.
Emanate's pitch is that speed decides industrial orders, since a buyer can usually get the same part from the next supplier. Nirghin says the software sits on top of the systems suppliers already use and pulls scattered account knowledge into one place, so accounts are handled "24/7, with a full knowledge graph of every single interaction." She avoids the labels chatbot and agent, calling the product an engine: "You point it at a business metric, and the engine figures out the rest."
The company has not opened its results to outside scrutiny. It won't name a customer, citing confidentiality agreements, and shared no revenue or margin figures. Emanate says it measures results by giving the engine to one group of accounts and comparing them with the rest of the business over the same stretch, with both groups facing the same prices and demand. In February it said the product could lift customer revenue by 60% to 80%. Nirghin also said the engine has never made a serious mistake, which it attributes to earned autonomy and careful deployment — a claim that, like the revenue figures, still rests on the company's own account.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
AWS said its chips business and AI business each passed a $25 billion annual revenue run rate in Q2 2026, both…

A Market Hang panel including Gerber Kawasaki CEO Ross Gerber, Yahoo Finance's Julie Hyman, Business Insider's…

On Semafor’s Future Proof, Reflection AI co-founder and CTO Ioannis Antonoglou said companies have to throw ou…

Square launched the Square Kiosk, a pill-shaped ordering device with a 21.5-inch touchscreen and an AI voice l…

OpenAI filed a document on Wednesday accusing an unnamed former employee of taking more than 100 files on its…

Apate.ai runs thousands of undercover AI agents that answer scam calls, reply to shady messages, follow phishi…
