This is the newsletter of Pluris, a network of pre-vetted applied AI experts and AI-native service organizations that specialize in helping businesses design and implement AI solutions that drive value.

Two free ways to use us. Browse the Pluris AI Use Case Library — over 150 published cases now, with about 170 more in the queue (329 in the dataset and growing) — for real AI implementations from our network: what got built, what it cost, and what changed once it shipped, organized around the problem you're solving. Or submit an AI need through an Expert Review: we'll anonymize it, match it to experts with relevant experience, and return 3–4 perspectives with pricing.

Two cases this week. Alex Gruebele and Chandler Gonzalez of Tau9 Labs built an instant-quote platform for an injection molding manufacturer that had been quoting by email, in a market where the shop that answers first tends to win the job. Quotes now run from part file to payment in seven clicks, and the platform pulled about $40,000 in its first month with no marketing behind it. Kenny Jen of QuantFi built an agentic month-end close on his own firm's books first, encoding the accounting judgment into a knowledge base before turning an agent loose on it, and cut the close from ten days to one.

We added more than 180 new cases to the library just this past week, from Fractional AI, CorralData, TrustEvals, Pluto AI, and dozens of others, with the rest getting cleared for publication. If you want to stand out in a noisy market, the best way to do it is through the proof of your work, and adding yours takes nothing on your end: we pull the case from your own site, reformat it to the library's standard, send it back for your approval, publish it, and fill in any gaps after. Reply and point me to your site, and we'll take it from there.

Key takeaways from this week's cases:

One thing I'm seeing more of: readers clicking straight through to talk to the expert behind a case. That's the point. If one of the cases below fits what you're working on, hit the link and I'll make the intro.

Order a custom part from a contract manufacturer, and you usually email for a quote, then wait. In injection molding, that wait costs real money, because the shop that answers first tends to win the job. Alex's rule of thumb is that quoting fast can roughly double your odds on a piece of business. The bigger players had figured this out and were taking the high-margin prototyping work with slick instant-quote platforms.

This injection molding manufacturer wanted into that faster, higher-margin segment, and the off-the-shelf 3D quoting tools didn't fit: built for a general case, loaded with features the shop didn't need, and slower for it. The ask to Tau9 was specific — get a customer from part file to swiped credit card in as few clicks as possible.

Tau9's founders come at manufacturing from both the hardware and software sides: Alex Gruebele did mechanical engineering and a hardware PhD, and Chandler Gonzalez built hyperscale software at Amazon. They built a custom platform around a proprietary C++ 3D modeling engine that reads an uploaded part file, runs manufacturability analysis, and prices the job in real time. On top sits a seven-click checkout, part upload to payment, with no email and no human in the loop. The engine was the hard part and most of the four-month build; the storefront on top took about two months, with AI coding tools speeding every stage.

The detail that shaped the build came from watching mechanical engineers use the prototype. The difference between a quote in ten seconds and one in a minute was the difference between a customer finishing and clicking off to a competitor's tab, so Tau9 pushed the speed down to the second. The platform pulled about $40,000 in new revenue in its first month with no marketing behind it, and the client has since expanded the contract. The bigger bet is that the quote tool becomes the front door: land the engineer at the prototype stage and keep them through production and the repeat, higher-volume orders that follow.

Here's where the AI actually shows up. There's no agent and no chatbot in the finished product; the edge came from AI coding tools that let a small team write a bespoke platform in four months, the kind of custom software that wasn't economical a few years ago. As Alex puts it, code is cheap now and getting cheaper, so the hard part is knowing what to build. That flips the build-versus-buy math: the off-the-shelf quoting tools were the only option before, and they never really fit, carrying features the shop didn't need and missing the niche it wanted. Now custom does exactly the job, and AI is what made it affordable to build.

Best fit for: Contract manufacturers and industrial job shops losing deals to faster-quoting competitors, particularly those serving custom-geometry or made-to-spec orders.

→ Want to talk to Alex and Chandler? Reply to this email and I'll make the intro. Or request one directly on the case study page.

Month-end close at most finance teams is manual and people-heavy, with no memory and no self-correction. Every month starts from zero: the same reports pulled, the same reconciliations chased, the same variance questions answered. The whole thing tends to live in one controller's head, which makes the close hostage to that person's bandwidth and leaves nothing behind for next time.

QuantFi decided to run that problem on its own books before selling the fix to anyone. The reasoning was that you can't automate judgment you haven't first written down, so phase one was the knowledge base, not the agent. They encoded their chart of accounts, entity structure, accounting policies, and prior-period close decisions into a structured, queryable layer, with the people who set those policies right there to check the work.

From there, they wired an integration layer into QuickBooks Online, then a Claude-based agent that pulls the trial balance, flags line-item variances against prior periods, drafts journal entries and variance commentary, and routes anything unusual for review. A human gate stays in by design: for auditability and SOX-adjacent control, the controller or CFO signs off before the books close. Every approved decision writes back into the knowledge base, so each close starts smarter than the last.

On QuantFi's own close, the cycle dropped from ten days to one, and the full build took under four weeks. The ten-to-one is what gets attention. The order of operations is the reason it held: because QuantFi codified the judgment against a process it already understood cold, the same architecture now ports to a client's books instead of starting from a blank page.

Best fit for: PE-backed portcos and middle-market CFOs running a monthly close with a lean team who want to cut the manual, repetitive steps without replacing their controller, especially companies already on cloud accounting (QBO, Sage Intacct) that haven't rebuilt the workflows on top of it.

→ Want to talk to Kenny? Reply to this email and I'll make the intro. Or request one directly on the case study page.

🎙️ Latest podcast: Why Most AI Spending Never Reaches the P&L. Jimmy Bijlani of AI Momentum Partners told me about a PE-backed software company that was sure it was ahead on AI, until his team went under the hood and found developers copy-pasting code by hand and a support bot answering almost nothing. The fix was one boring internal workflow, built on software the company already owned, worth about $700K over three years. → Website | Spotify

Working on something where one of these patterns might fit? Reply to this email — I'll either make the intro to the right expert directly, or open an Expert Review and come back with 4-5 expert perspectives and pricing in a few days. Both are free, and I read every reply.

— Stu

Proof of Work: Applied AI Case Studies. The newsletter of Pluris.