40 ways operators are using AI right now
Filter by industry, by type of engagement, or browse the whole library. Click any card to see the full story.
These are anonymized on purpose. Selected engagements are published in full, with named clients, at signaladvisory.ai/case-studies.
How a solo CPA got 10+ hours back every week
She was starting her day at 4:30 AM just to keep up with email, client onboarding, and admin work. We built AI workflows that sort her inbox, prep her week, and set up new clients automatically.
See the build
How a construction estimator cut quote time from 4 hours to one click
The owner was doing 1,500 to 2,000 estimates a year by hand. We built a Chrome extension that lives inside his existing tool: open a job, click a button, get a finished proposal.
See the build
How an oil and gas leadership team went from skeptics to a funded AI pilot plan in one day
A full-day Fort Worth workshop moved a 15-person leadership team from "AI is for tech companies" to a ranked list of real pilots they could fund next quarter.
See the build
How a solo attorney got her time back without losing her voice
She was drowning in coordination work and her AI drafts did not sound like her. We built a digital chief of staff that preps every meeting, writes replies in her voice, and tracks every commitment made on calls.
See the build
How a 35-person CRE firm turned Sunday bank reconciliations into a 4-minute task
The accounting controller was spending her Sundays reconciling six entities in Excel. One live demo changed the conversation. We mapped the highest-value workflows in a half-day onsite session.
See the build
How a concierge medicine practice stopped losing leads from Instagram
The practice had steady inbound but the follow-through was breaking. A three-day discovery visit found every leak. The fix focused on two workflows responsible for most of the revenue lost.
See the build
How a capital markets BD partner turned every meeting recording into six deliverables
He had no executive assistant and worked at a firm where most consumer AI tools could not pass IT review. We built a system that turned every meeting recording into action items, follow-up drafts, LinkedIn content, and deal intel automatically.
See the build
A LinkedIn ghostwriting agency stopped losing clients who couldn't see it was working
Clients were getting real results but couldn't trace them back to the agency. One almost canceled. We built a reporting agent that surfaces wins from their inbox and a client hub that makes the work visible.
See the build
How a 75-campus school network fixed its enrollment funnel without buying more leads
The network had enough leads. They were going cold before anyone followed up. We built an AI agent that responds to every new inquiry within a minute with a message tailored to that family, proved it on a four-campus pilot, and rolled it out to every school in the network.
See the build
A venture studio CEO can now ask questions about his whole portfolio from his phone
A venture studio CEO ran five business lines across dozens of tools and accounts. He didn't want dashboards. We built AI agents that bring the answers to him via iMessage, Slack, and his inbox.
See the build
A matchmaker found out which AI tool is actually best at finding people online
Five tools, 69 real candidates, one question: which AI finds the right LinkedIn, Instagram, and Facebook profiles? The answer was not one winner. It was a ranked pipeline that uses each tool where it performs best.
See the build
A rare coin collector built his own dealer-tracking system in one afternoon
In a single coaching session, we built a dealer intelligence report and started a web app that watches for target coins and alerts him when one appears for sale.
See the build
A VC partner got an AI that handles his inbox, deal flow, and follow-ups
A VC partner was spending too much time on email, follow-ups, and LinkedIn. We built an AI chief of staff that handles triage, drafts, and deal-flow tracking, plus a pilot for searching 40 years of legal documents.
See the build
A venture studio built a solid knowledge base so its AI agents would actually work
A first AI executive assistant failed because there was no organized knowledge underneath it. This time, we built the knowledge base first: every LP, fund, portfolio company, and investment decision captured before adding a single agent.
See the build
A salon and spa conference left with AI tools they could use on Monday
A 75-minute workshop for salon and spa owners at a national industry conference. The focus was revenue and retention, not tech. Attendees left with live demos and exact prompts to use that week.
See the build
A jewelry brand owner now runs her own photo shoots from a laptop
A one-session coaching for a fine jewelry brand owner who needed product imagery without photo shoots. She learned to generate on-model photos, lifestyle shots, and co-branded mockups, all from her laptop.
See the build
A BigLaw partner cut through 200 daily emails using AI tools the firm actually allows
A 1,000-lawyer Australian firm restricted AI on client data to a short approved list. We mapped what was allowed, found where the real time went, and designed a Digital Chief of Staff that works inside those limits.
See the build
A two-GP VC firm saw exactly what a Digital Chief of Staff would save them each week
A two-GP venture firm with 6 part-time venture partners had no ops automation. We built a full Digital Chief of Staff demo around their Monday-morning workflow to show exactly what AI would take off their plates.
See the build
How an IT firm stopped losing leads to slow prospecting and ticket chaos
Sales reps could not monitor hundreds of online buying signals at once. Support techs were sorting tickets by hand. We mapped out two systems: one that finds hot prospects before competitors do, and one that delivers tickets with the right checklist already attached.
See the build
How a commercial real estate broker built outreach automation that IT could not block
His firm's IT department blocked all third-party email tools. So we built his outreach and document workflow on the side of the system IT did not control: his personal Mac Mini, LinkedIn, and Telegram.
See the build
How a 1,000-person SaaS company figured out if switching to Claude was worth it
A 60-person partnerships team was paying for a Gemini contract that wasn't working. We ran a discovery session and scoped a workshop to show them exactly what Claude could do for their real workflows.
See the build
How a think tank ran its $50M fundraise without burning out a small team
A global think tank needed to raise $50M, run salons across three cities, and recruit fellows. The team was small. AI handled donor research, event logistics, and recruiting so people could focus on the decisions only humans can make.
See the build
Fourteen leaders left a two-day offsite with a real AI roadmap and a ranked list of what to build first
Fourteen office heads and operating officers at a $174B firm came in skeptical and left aligned. Two days. Real firm data in the demos. A pre-work system that turned 100-plus employee pain points into the agenda.
See the build
How a private members club gave every host the knowledge only the founder used to have
The founder was the only person who could answer brand, member, and strategy questions. We built a knowledge base, a prospecting pipeline, and auto-generated event tools so the whole team could operate without going through him.
See the build
How a logistics operator got an always-on AI assistant he owns and controls
The operator needed AI that fit into his daily workflow without adding new apps or vendor lock-in. We installed OpenClaw on a Mac Mini and connected it to the tools he already uses.
See the build
A speaker set up an AI that manages his network so he doesn't have to
He meets hundreds of people a year but had no system to keep those relationships alive. We built a private, always-on AI assistant that watches his inbox, drafts follow-ups in his voice, and flags who he hasn't talked to in too long.
See the build
An operator with 1,000 contacts now gets 5 people to reach out to every morning
He had a big network but no system. People fell through the cracks. Promises went unkept. We built a daily Slack tool that picks the 5 most important people to reach today and hands him a draft message ready to send.
See the build
A sales agency got a working AI product it can sell to every CPA client
The agency serves accounting firms but had no AI product to offer them. We ran a workshop that built a real agent live on the call. The founder left knowing how to build more on his own.
See the build
A multi-entity venture firm put AI in its inbox, meetings, and books without autopilot
The founder's inbox was manual, institutional memory lived in people's heads, and the bookkeeper reconciled card charges across entities by hand. We shipped a working stack in under three months on one rule: AI drafts and ranks, a human approves.
See the build
A members club pre-sold millions in memberships for a building that existed only as blueprints
You cannot photograph a club that has not been built. Working from architecture renderings and 2D blueprints, we produced photorealistic imagery and a marketing video of the finished space, and the pre-sale campaign they powered generated millions in membership revenue before the doors opened.
See the build
A private club qualified thousands of prospective members for less than 16 cents each
Recruiting thousands of founding members means researching thousands of names. Done traditionally, that qualification can run 50 dollars a lead and take a researcher months. An AI pipeline did it for under 16 cents a lead and returned a ranked, queryable list of who to invite first.
See the build
Every guest at a 60-person dinner got three personal introductions that read like the host wrote them
The best moment at a members dinner is being told exactly who you should meet. That used to require a senior host with the whole room in their head. An AI system now drafts three tailored introductions for every guest, and the host reviews and sends every one.
See the build
A venture fund built one AI-native platform to replace its doc-sharing, market intel, and CRM stack
A venture studio was paying for a stack of disconnected SaaS tools to run its fundraise: a doc-sharing service, a market intelligence subscription, a CRM nobody loved. We built them one platform that does all of it, with AI underneath, that the fund owns outright.
See the build
A payments company got a private AI that respects who is allowed to see what
The blocker was not AI capability, it was permissions. Every off-the-shelf option used shared service accounts that gave everyone access to everything. We built a private, on-premise AI where every query carries the asking user's identity, access is deny-by-default, and every answer is auditable.
See the build
An estimating engine learned the difference between confident and correct
Version one drafted quotes fast and was confidently wrong on messy jobs. Version two was rebuilt around an evaluation harness: validated against 142 real historical jobs, taught the estimator's conventions across seven scored rounds, and gated so it only auto-drafts when the evidence is solid. When it is not sure, it says so.
See the build
A CPA firm owner went from using AI to building it, then started teaching her industry
The first engagement gave her hours back. The second changed what she sells. Weekly coaching took a non-technical firm owner up the ladder from doing tasks in AI, to packaging them as reusable skills, to running her own agents, and her AI fluency is now a revenue line of its own.
See the build
An investment firm is building an Office of the CEO that runs on shared AI context
The CEO's leverage problem was not effort, it was context: his EA, his chief of staff, and his tools each held a fragment. Weekly enablement made the team AI-fluent first, and the early number is real: board-grade slide decks that took four to five hours now take about twenty minutes.
See the build
A capital raiser left a workshop and built an outreach system that booked 20+ investor meetings
The best workshop outcome is what attendees build after you leave. A capital raiser launching a credit fund took the AI bootcamp, went home, and built his own investor-outreach system over a database of more than 3,000 investors. It has generated more than twenty investor meetings.
See the build
A commercial real estate MD went from AI skeptic to sending AI-drafted investor letters in three sessions
He had passed on AI once and called back to say so. Three coaching sessions later, a colleague's two-paragraph draft became a formatted, investor-ready project letter in under five minutes, and he printed it and texted the colleague on the spot. Sessions are expanding to his firm's teams.
See the build
A salon team came back from an AI workshop and shipped a training manual, dispute letters, and a podcast
The test of a team workshop is what the team does the month after. A salon owner brought her staff to an AI workshop; by the follow-up session the team had produced a complete front-of-house training manual, a credit-card dispute argument that worked, marketing images, and an internal podcast, all self-served.
See the build
