Land your dream AI sales job. Mistral €1.7B

👋 What’s up, pipeline punks —

This is Prompt Punk — the best AI sales newsletter in the Milky Way! 😉

TLDR; inside today’s newsletter

Let’s rip…

What is sales?

To me, sales is the process of building trust so people want to exchange money for value.

How does AI change sales?

AI makes us faster, lets us carry more load, and skip the boring stuff.

But does it help earn trust? Yes—if you use it deliberately.

When I was learning to sell, I picked up a 9-point framework I still use to earn trust.

To win customers I strive to be:

How can AI help us speedrun this? Here’s what I’d do:

AI can help you earn trust a lot faster.

Now back to a regular scheduled programming…

Salesforce cut ~4,000 customer service roles after rolling out Agentforce. Half of all support interactions are now handled by AI. Some staff redeployed; backfilling support roles largely paused.

“I need less heads,” says the CEO.

The results

How it works
Salesforce operationalized agents into workflows, primarily answering support tickets. Low-value work moves to bots while humans handle edge cases.

Why it matters (winners & losers)

SMBs aren’t “experimenting” with AI—they’re inhaling it.

This is according to Goldman Sach’s 10,000 Small Business Survey.

68% already use AI and another 9% are starting this year, up from 51% two years ago. And it’s not about robots replacing humans: 80% say AI enhances their team.

But deciding to “do AI” isn’t the hard part, it’s implementation. The blockers are painfully human:

While most sellers are chasing head- banging enterprise AI transformations. SMBs are eating it up, with little support.

SMBs use AI for research, communication, operations, sales, marketing and more.

Why it matters:
SMBs make this AI transformation sustainable. NVIDIA and OpenAI employees can’t be the only winners. Great to see the “little guy” get some wins.

What it does: Visual workflow builder that stitches your stack together—apps, APIs, databases, webhooks, even LLM calls. Self-host it or use their cloud.

Example: Pipe your sales leads from forms → enrich → score → CRM → Email → auto-followups, no duct tape required.

See this brilliant tutorial.

Why it’s valuable: Automate the boring, standardize the smart.

Website: n8n.io

What it does: Connects to all your knowledge (Docs, Slack, Drive, CRM), then lets you chat, search, and ship with AI agents that do the work.

Update Salesforce, prep RFPs, onboard newbies, the lot. Think enterprise search + doer bots.

Why it’s valuable: No more scavenger hunts for information. Individuals get answers in seconds, managers stop babysitting Slack. The whole organization moves faster.

Customers report big time wins (e.g., 10 hours/week saved or >50% faster on heavy workflows).

Website: sanalabs.com

Where are the killer AI sales jobs? The ones where you crush quota, bond with customers and stack equity like meetings on Mondays?

I put together some of the more obvious opportunities here, but how do you find the next Lovable?

Here’s my formula: find a rocketship early—sub-$100M raised—with a clear path to oligopoly.

You want tomorrow’s category king before the bros and bankers show up.

Forget infra startups (chips and models). That train has left the station—unless Ilya Sutskever personally knights you as Seller #1 at SSI.

Unlikely.

So the move is apps. Hunt for teams using LLMs in a novel, money-moving way.

Stuff that helps people get paid, save time, or laid (Hello JuicyChat! NSFW).

How to spot them:

When to jump? Series A is plenty.

If the problem is real and the product pops, you don’t need a 40-page Gartner report to pull the trigger.

You need courage.

CoreWeave, the AI cloud hyperscaler launched CoreWeave Ventures, writing checks and cutting compute-for-equity deals.

Early partner: Moonvalley (AI filmaking).

Translation: they don’t just rent GPUs anymore—they’re buying upside in the apps that need those GPUs.

If an investment takes off, CoreWeave wins twice: usage + equity. That’s picks-and-shovels… with royalties. Smart.

Why it matters:
This turns the “neocloud” upstart into a potential kingmaker. Founders now get capital + capacity in one term sheet, increasing competition with the hyperscalers.

Europe just put real fuel in the tank: ASML led a €1.7B investment into Mistral, valuing the French AI lab at €11.7B. It’s a boat load of cash—and a serious partner.

It probably didn’t hurt the CEO of ASML is French and a former French finance minister is an advisor at ASML. 😉

Who is ASML? Duh? The company that makes the machines that print the world’s most advanced chips. Think iPhones, data centers, and the GPUs training your favourite models.

No ASML, no cutting-edge chips.

Mistral says the cash fuels frontier research; ASML says it’s a strategic partnership (read: models inside EUV workflows printing billions).

Why it matters:
Europe is finally building its AI-industrial flywheel to the tune of billions: capital + compute + talent. I can’t wait to see what happens next.

Microsoft is prepping a new Copilot (AI in Microsoft Office) powered by Anthropic’s Claude with access routed through AWS.

AI creates strange bedfellows.

Internal tests reportedly found Claude stronger at Excel financial automations and presentation-ready PowerPoint generation.

That probably explains the success of Gamma (AI presentations) and Paradigm (AI spreadsheets).

Why it matters:
This isn’t just about getting access to a “better quality” model. It’s Microsoft hedging against OpenAI.

What is Exa? Think “Google, but built for AIs, not humans.”

Exa powers search for machines. So when you look up real time data in your favourite LLM or AI app, you get a clean result, not SEO optimized garbage.

With new funding, Exa will scale indexing, grow its GPU fleet 5x, and chase “perfect search. Organizing the world’s information for AIs first, humans second.

Why it matters:
Every serious AI workflow now needs real-time web data. Wrapping in Google won’t cut it.

The apps that win won’t “browse the web”; they’ll plug into a purpose-built index that’s faster, cleaner, and customizable.

Heard of “996”? That’s the shorthand for tough (unsustainable?) work regime of Chinese tech companies. Working 9am to 9pm, 6 days a week.

Apparently it’s now in vogue in San Francisco (even London), as companies chase AI gold.

Ramp, the expenses app, have spotted increased Saturday meal expenses in their transactions.

Prompt Punk POV: If the prize is big, why not work hard!

Thanks for reading! ✊

If you liked this:

📨 Forward to a biz techie friend
💬 Hit reply and tell me what you want more (or less) of
📊Vote in the poll below 👇

— John
Prompt Punk