# AI Boardroom — Clean Transcript
**Date:** 2026-07-10 · **Type:** AI Boardroom
**Location:** Industrious East 7th, Austin, TX — Conf. Room A
**Duration captured:** ~3:15:00

> **Timestamp note:** times below are from the RAW recording. They will be re-mapped to the published/edited video before any "watch at" links go live.
> This transcript has been lightly cleaned (dead air, AV setup, and crosstalk removed) and redacted to protect anyone's private or sensitive information. Nothing here should be treated as a complete word-for-word record — use it to jog your memory and feed your own AI, not as a legal record of the meeting.

## Roster (who was in the room)
| Name | Business | Note |
|---|---|---|
| Austin Distel | Agentic Society | Host / Facilitator |
| Nick Blancett | Creative Director (real estate/construction niche) | |
| Prashant Vanka | Real estate development | Remote by phone; had to leave for a lot closing |
| Cristian | Agentic Society / social media clients | Team |
| Rohan | Frontier Studio (content/authority) | |
| Darby Rollins | GenAI University / Agentic OS | Team |
| Lauren Goldstein | The Biz Doctor (fractional COO) | New member |
| Chance Forman | Lilybug (new agentic MEP company) | New member |
| Will Sayasenh | Video / Photography | New team member, just relocated to Austin |

---

## Transcript (attributed · timestamped)

### [00:21:27] Opening — Austin Distel
Austin opened by resetting the start time going forward: future events start at 9:30, with members arriving at 9:15 to settle in. He welcomed the newest members by name — **Lauren** and **Chance Forman** — and then gave a shoutout to **Will Sayasenh**, the newest Agentic Society team member, who had driven in from out of state and arrived at ~2am the night before to start this morning. Austin explained Will's role: capturing content clips for both social and the member portal, and being the person to talk to if a member has a success story worth telling. He tied this to the broader content engine — Rohan's LinkedIn work packaging member wins into case studies — and cited a recent example (a member referred to as "Oscar") who started getting inbound DMs and deals directly from being spotlighted in a post.

Austin then explained today's format for the new members: Roses, Thorns, and an Ask (about 45 minutes), followed by a Roadmap-building exercise — an evolved version of one he ran ~45 days ago to identify the single highest-leverage two-week sprint — and, time permitting, open Show & Tell.

### [00:29:31] Roses, Thorns & Ask — Austin Distel
**Rose:** Austin onboarded and trained a new manager for his spa business who came in with real ownership experience (she'd previously owned the same franchise concept in New Jersey but decided physical-location ownership wasn't for her). He described the handoff as a chance to teach her Claude in a team workspace so she can query company docs directly — he spent two days in Dallas onboarding her, using Claude on his phone throughout the day to capture every Q&A and feed it into a company brain, on top of uploading the franchise's existing Trainual SOPs.

**Thorn:** Agentic Society's physical room is at capacity — last Friday had a 16-person waitlist. Austin is weighing the options: more events, gatekeeping/limiting repeat attendance, a bigger venue, or a genuinely good virtual format (not just a Zoom simulcast, but something with real interactivity — "the Hormozi style with the big screen," members able to get involved, not just watch). One driver: onboarding remote employees (he used the example of his own studio manager in Dallas) into using Claude day-to-day without needing to be in the room.

**Ask:** Ideas for what a high-value virtual version of Agentic Society could look like, especially for scaling employee onboarding.

### [00:37:15] Roses, Thorns & Ask — Nick Blancett
**Rose:** Nick got the Meta MCP fully working with Claude and launched his first ad campaign — 18 leads in two days. Bigger win: he's built what's now a near-plug-and-play Premiere Pro editing pipeline for turning long interview footage into finished selects. He described the technical path in detail — using an indexing model to define and label each clip and its content, storing that per-clip metadata, then feeding the clip through OpenCV using Lucas-Kanade optical flow to measure camera movement/shake and flag whether a take is usable — all piped into Premiere via an MCP connection. On a real 45-minute talking-head interview, this took an interview down to 7.5 minutes in about an hour of his time (versus what he estimated as ~8 hours manually). He's still refining how to extend the same approach beyond single-subject interviews to full events.

**Thorn:** Turning still images into believable video ("the shower generator" — his term for the still-to-video pipeline used for one of his home-services clients) is expensive ($7–12/reel) and inconsistent in quality, and he hasn't found an efficient way to storyboard multiple images into one coherent, dynamic clip.

**Ask:** Prashant recommended he talk to **Riley Spiller**, a community connection who's had good results prompting still-image storyboards into video.

### [00:46:12] Roses, Thorns & Ask — Prashant Vanka (remote)
**Rose:** Prashant brought on a new sales partner with a military background plus two new sales reps out of a pool of ~30 candidates interviewed. On the development side, his team started building homes again after a long pause, and the first new-build sold — the first sale in that line of business in about four years.

**Thorn:** His original OpenClaw setup was "vibe-coded" without proper governance — agents weren't cleanly siloed by task, so context and memory decayed over months of use. He's since had Fable help him rebuild the workflows on **Hermes** instead, partly because Hermes has a built-in Kanban board he now uses to run his day. His company brain and SOPs aren't fully connected yet — he described a recent case where a potential investor sent 13 diligence questions and, if his system were properly wired, it should have drafted ~90% of the answers; instead he answered all of them manually.

**Ask:** Help getting his company brain properly structured — he named this as the real "step one" before the rest of his agent workflows can be trusted to run without decaying. Austin noted Agentic Society is actively scoping a "company brain" recommendation/prescription (Google Drive–based structure) and offered to collaborate with him as an early test case. Darby also chimed in that Isaiah and CT have systems worth looking at as reference points.

*(A side conversation about Prashant's investor renderings for a project he's invested in was shared explicitly off-the-record and is not included here.)*

### [00:56:28] Roses, Thorns & Ask — Cristian
**Rose:** Built a workflow to turn Agentic Society's mastermind calls and mixers into clipped, captioned reels for Instagram automatically, using Rohan's tool **Bitter Clip** plus an additional layer for B-roll. On the client side, he landed two new clients, bringing him to three total in about a month and a half of doing this work.

**Thorn:** Not satisfied with the automated B-roll selection or caption quality — still has to manually pick B-roll for a lot of clips. Managing time between Agentic Society content work and client fulfillment (research/ideation for clients' social) is also a live juggling act.

### [00:59:07] Roses, Thorns & Ask — Rohan
**Rose (two-part):** First, Rohan is finding that clients want more than AI-driven efficiency — they want help with narrative and storytelling (website messaging, testimonials, "what's the big outcome, what are the proof points"), and he's been able to upsell this by mining his own customer-interview process for raw material. Second, the member-spotlight LinkedIn content strategy for Agentic Society has generated 10 inbound DMs in four weeks, including a senior marketing director at IBM and a chapter lead from an invitation-only high-net-worth community, both drawn in by the content.

**Thorn:** Struggling to find a good image-generation tool to visualize the concepts in his head (e.g., explaining what share of a platform's audience is actually creating worthwhile content) — he's tried a couple of tools but hasn't found the right fit yet.

**Ask:** Members to keep sharing success stories so they can be turned into case studies.

### [01:09:20] Roses, Thorns & Ask — Darby Rollins
**Rose:** Shipped the Agentic OS member portal — members can now log in and access the MCP directly. Darby is happy with how the UI turned out and is now shifting to onboarding people into it. In parallel, he's been refining the "company brain" onboarding protocol based on research into what's working elsewhere, informed by his own second brain build over the last six months, and connected with Roman about aligning Agentic Society's own internal company brain to the same standard so new members get a consistent starting template.

**Thorn:** Running multiple AI workflows at once (e.g. video-editing pipelines alongside other agent work) is starting to bottleneck his current machine — he's considering a dedicated, beefier laptop just for the heavier workloads.

**Ask:** Get members using Agentic OS and send feedback — Darby shared the onboarding link in Slack during the session.

### [01:15:46] Roses, Thorns & Ask — Lauren Goldstein, The Biz Doctor
Lauren introduced herself and her business: she works with service-based business owners doing $5–50M in revenue who are still stuck operating their business rather than owning it — helping them build the right team, operations, and (increasingly) AI systems to get their freedom back. She's been doing this for 15 years and also runs **The Biz Doctor** podcast (100+ episodes).

**Rose:** Through a fractional-COO engagement, she diagnosed a client's operations and found significant avoidable waste across SaaS subscriptions, licenses, and tools they weren't fully using or needed — including moving them off an expensive per-month video-calling setup they were using for basic call routing onto a far cheaper annual alternative, plus a meaningful payroll-provider switch. *(Specific dollar figures from this story have been held back — see the internal build copy for details pending the client's/Lauren's confirmation they're shareable.)* Her broader point: things that look obvious to an outside operator are often invisible to the business owner living inside them every day.

**Thorn:** Her own podcast production stalled after she parted ways with her editor — recording, editing, and posting all fell on her and her team, so it's not getting done consistently. She's been doing a "yap challenge" — recording short videos in the car and posting them directly — as a way to get content out without the production bottleneck, and said it's helped her realize this doesn't have to be as hard as she was making it.

**Ask:** She has five open speaking slots and is looking for introductions to groups, conferences, or panels where the audience matches her ideal client profile (5–50M service, construction, design, financial services, or real estate businesses).

### [01:29:47] Roses, Thorns & Ask — Chance Forman, Lilybug

**Rose:** Finished building a custom CEO dashboard for managing his portfolio of businesses — it pulls live KPIs and cash-on-hand per company, has a built-in AI "coach" (running both Claude and ChatGPT so he can toggle), and a Kanban-style project board where his AI agents and a human chief of staff work side by side. Agents build their own checklists for each project, work autonomously, and flag anything blocked by a human directly onto that person's priority list.

The bigger news: he's launching a new **agentic mechanical/electrical/plumbing (MEP) company — Lilybug** — with partners who have had multiple nine-figure exits, aiming to become the most agentic company in the trades and scale toward a $500M business over time, starting in Texas with plans to expand regionally and roll up other operators in ~5 years. He wants the AI-forward approach visible from day one — e.g., in-field crews presenting instant pricing and financing on iPads the way modern software platforms increasingly support — and has already generated ~300 marketing video assets for the not-yet-launched brand using Higgsfield, at a fraction of traditional production cost.

**Thorn:** He's decided he wants to focus fully on Lilybug rather than split attention with his other companies.

**Ask:** He's putting $50,000 behind the initial marketing push for Lilybug and is looking for help from the room on how to optimize that ad spend to compete with established, well-funded incumbents in the trades and ramp up profitably from day one. Prashant, Nick, and others offered to connect given their experience with construction/service-business marketing.

### [01:38:12] Roses, Thorns & Ask — Will Sayasenh
Will introduced himself as the newest addition to the Agentic Society team, focused on video and photography.

**Rose:** After trying to make the move to Austin happen for roughly three years, he finally relocated — arriving in the early hours of the morning of this Boardroom. He's leaving a corporate job he didn't enjoy to do his own creative/content work full-time again, and already has a few clients in the home-services space (notably roofers), running ads and producing video content for them.

**Thorn:** Getting AI-assisted video editing to capture emotional nuance and tonality is hard, especially with off-script, unscripted footage — he still has to go back in and hand-edit for the emotional beats, even though the process saves him real time overall.

**Ask:** He's looking for a rental in the ~$1,200/month range (a one-bedroom or similar) to get settled, and is looking to connect with more people in the home-services / roofing space. Several members offered local rental-agent contacts and neighborhood recommendations.

---

Austin closed the Roses & Thorns portion by tying several threads together: the room's differentiation, as more businesses race to sound the same using AI, will come from members getting better at telling their own story and leaning into what's authentically human about their businesses — not just chasing efficiency.

### [01:57:31] AI Roadmap Exercise — Austin Distel (facilitator)
Austin introduced a workshop he's run quarterly with past teams (including at Jasper), now rebuilt as a live tool at **agenticsociety.com/roadmap**, built with Claude/Fable specifically for the community. The goal: leave with exactly one thing to build over the next two-week sprint — not a little improvement, a meaningful one — ideally the "domino" that makes everything else easier or unnecessary. Framing note: today is July 10, the start of Q3, and Austin is moving Agentic Society itself onto a two-week sprint cadence tied to these Boardrooms.

**The four-step process:**
1. **Problem Dump** — brain-dump every bottleneck, constraint, or pain point in the business (prompted by questions like: what's repetitive, manual, expensive, or slow; what would break under 10x demand; what's the next role you'd hire and why).
2. **Pain/Frequency Matrix** — plot each problem by how often it happens (daily → quarterly) and how painful it is, to find the highest-value problem to solve first. Austin's framing: this is also how you'd evaluate a new product or business idea — solve the most frequent, most painful problem, because that's where the money is.
3. **Solution Brainstorm** — once the top problem is chosen, generate every possible way to solve it (hire, buy software, build it), without judging yet.
4. **Effort/Impact Matrix** — plot the solution options by effort (cost, time, headcount) versus impact, to find the "quick win" (low effort, high impact) rather than the "big bet" (high effort, high impact) or the "avoid" (low impact regardless of effort).

Austin walked his own example live: his front desk currently texts every incoming ad lead manually because his corporate marketing agency only delivers leads into a spreadsheet — a daily, high-frequency pain given the well-documented dropoff in conversion after the first hour of no response. He compared three solution paths (GoHighLevel automation with A2P texting, hiring/training a dedicated salesperson, or building a durable AI SOP agent) and walked through why the *first domino* — getting GoHighLevel's A2P registration done — is the quick win that unlocks the more ambitious agent build later, because it gives any future agent access to the CRM and lead history. He also shared a cautionary note from experience: don't run high-volume automated texting through your primary/public-facing business number, since carriers will flag and can force a number change (which cost him real money and hassle once before).

Each member then worked through the tool live on their own laptops/phones — problem dump, plot on the pain/frequency matrix, brainstorm solutions, plot on the effort/impact matrix, then select and lock in exactly one project for the two-week sprint, ending with a downloadable markdown build plan to hand to their own AI.

### [02:53:19] Two-Week Sprint Commitments (shared around the room)
- **Austin** — Register A2P for GoHighLevel to unlock automated, always-on lead follow-up.
- **Lauren** — Build a lightweight, duct-tape AI helper (based on her recorded client calls, chatting with her through Slack) to help her track client accountability in her fractional COO work — an interim solution while she moves her operation off Microsoft.
- **Chance Forman** — His actual bottleneck turned out to be non-technical: he needs a backup contractor's license for Lilybug, so his sprint is manual relationship-building — visiting local supply houses and building referral relationships rather than an automation.
- **Cristian** — Standardize on a small set of proven content formats/scripts that already work across niches for his clients, instead of over-engineering something custom for each one.
- **Rohan** — Build an AI "thought partner" that interviews him with the same style of questions he uses with his own clients, to help him extract and publish his own content and updated lead magnets more consistently.
- **Nick** — Build a "prompt engineer" style agent that runs ongoing research/testing on more efficient, higher-quality ways to generate video from still images, since his current approach is manual trial-and-error.
- **Prashant** — Rebuild his company's second brain and connectors properly (he described this as "Closer OS" — a system that listened to sales calls, summarized them, and recommended next steps, which broke down over time) — with the company-brain foundation fixed first so the rest of the agent stack stops decaying. He and Austin agreed to collaborate closely on this over the next two weeks.
- **Will Sayasenh** — Focus on applying AI automation to his own business and his existing clients now that he's settled in Austin, and follow up with the rental-agent referral he received earlier in the session.

Austin closed by reflecting that the framework scales — from a whole-portfolio view down to a single role's task list — and that the group would hold each other accountable to these single commitments at the next Boardroom.