Andrew Das

In January, my co-founder (@hardhatCFO) and I started Fervent from ZERO. 12 months later: $421k in revenue for FY-2025. Peak run-rate over $810k. 1 acquisition closed. STRAP IN: for the backstory, the pivot, and a thread on why the: "you can (and should) just do things" isn’t just a meme 🧵

@toddsaunders Not sure if we count but would love to chat — accounting services for trades but we mainly serve 3 verticals HVAC Electrical Basement / insulation teamfervent.com

I’m becoming obsessed with vertical specific agencies. Think digital marketing for landscapers or website agencies for plumbers. Would love to chat with folks running these. DMs open.

Anyone else find it jarring when you see something on the For You page that isn’t Claude / AI content It’s like Claude code Claude skill Claude code GPT Whoa ! Whoa ! A real post Claude code Claude code Claude skill Back to doomscrolling

@OneManLBO @patrickdichter Actively getting it spun up so no data points yet but after running cold calls with an agency for 6 months and getting zero — our bar is low The cold email gurus would tell @patrickdichter the zero results likely mean the copy / offer wasn’t good enough 🤷‍♂️

@patrickdichter @TrueMargin any data points from your end?

Fool me three times, can’t get fooled again. We tried cold email marketing again for a 3rd time with zero results. 👍

@appenz @meetgranola If you just auto sync the granola notes to obsidian and include the sync with the transcript then all the same data lives in obsidian in Md files and you can call that via API

Sorry to see Granola @meetgranola going closed. They encrypted their local db, no local and no cloud API. In a world where notes are managed by agents, the app now has zero value. Any recommendations for good alternatives? What are you switching to?

@jspujji AI

I recently spoke to a marketer who ran a $40M brand with just two designers and ONE AI process. I paid him 6 figs to build these systems for my companies. He chains together 7 AI tools: creative brief → image gen → scale winning assets. All run by 2 offshore designers. I’m giving away his entire operating system for free. Comment “AI” and I’ll send it.

@jack_c947 Claude

SHOCKING: 99% of GTM engineers using Claude Code are barely scratching the surface. Right now, the entire internet is screaming "Claude Code, Claude Code, Claude Code"... But here's the truth: running one-off prompts in a terminal won't build GTM infrastructure. To unlock its real power, you need to master: - The WAT framework and CLAUDE.md self-improvement loop - Sub-agents, agent teams, and deployed automations running 24/7 - MCP connections, browser automation, and GitHub worktrees I spent 100+ hours building and documenting the exact system and compiled every workflow, build sequence, and deployment guide into one complete playbook. I'll give it to only 4,500 people. To get it: 1. Follow me MUST (so I can DM) 2. Comment "CLAUDE" 3. I'll DM you the document If you don't follow or comment, you won't receive it.

@mikefutia NANO

Claude Code + Nano Banana 2 is f*cking cracked 🤯 I built a system inside Claude Code that researches any brand, writes 40 ad prompts from scratch, and fires them all to Nano Banana 2. One brand name + one URL = 40 production-ready static ads. All inside Claude Code. I took @alexgoughcooper's brilliant framework and automated the whole thing inside Claude Code. Perfect for DTC brands and agencies who need high-volume ad creative without briefing a designer or spending hours in Canva. If you're finding winning ad concepts on Meta and manually recreating them one at a time in Higgsfield — copying prompts, pasting product details, tweaking aspect ratios, downloading, organizing... This system eliminates the entire loop: → Give Claude a brand name and URL → It researches the brand's fonts, colors, packaging, and photography style → Builds a Brand DNA document from scratch → Fills in Alex's 40 proven ad templates (headline, us vs them, testimonial, UGC, review cards, stat callouts) with brand-specific details → Fires every prompt to Nano Banana 2 with your product photos as reference → Downloads finished ads into organized folders with an HTML gallery No Higgsfield. No manual prompt filling. No copy-pasting between tools. What you get: → 40 ad formats filled with your exact brand colors, fonts, and copy → 4 variations per format so you pick the best output → Product photos passed as reference so the model matches your real packaging → A reusable system — new brand, new folder, same pipeline Built 100% in Claude Code with Nano Banana 2. I put together a full playbook & Loom video showing the exact process to set this up yourself. Want access for free? > Like this post > Comment "NANO" And I'll send it over (must be following so I can DM)

Mr. Claude Code told me to make a business instagram, so I of course listened: instagram.com/teamfervent/ If anyone has good ideas for video content for an accounting & CFO services firm - let me know. Current winner: 4-hours of uncut excel footage. If we get 1.25 likes on this post, I will make it. May need a tutorial on social from @jamesonhaslam

This is likely record setting pace for any “independent sponsor” deal entry / exit growth. Indy sponsor in quotes cus I’m sure some will debate what to call GGG — having the big equity slug upfront to get the deals done certainly helps. Nonetheless impressive — would have to guess this traded in the mid to high teens

@TheMattBerman @openclaw @vishalojha_me @Meta ADS

I run my meta ads with @openclaw for $0/month 😱 here's the system that runs autonomously: step 1: daily health check → social-cli (major shoutout to @vishalojha_me) wraps @Meta's marketing API (token refresh, pagination, rate limits all handled) → am I on track? what's running? who's winning? who's bleeding? any fatigue? → the same 5 questions I asked Ads Manager every morning for 20 years step 2: catch dying ads before CPA spikes → @OpenClaw pulls daily frequency by ad → frequency > 3.5 = audience is cooked, CTR is about to drop → this one signal saves more money than any dashboard step 3: auto-pause bleeders + shift budget to winners → CPA > 2.5x target for 48hrs? auto-pause. no hesitation. → ranks every campaign by efficiency. recommends shifting spend. → last fri it paused an $87 CPA campaign at 3am and scaled my best performer 30% step 4: write new ad copy from your winners → agent analyzes what's working (hooks, angles, CTAs) → generates variations based on the patterns in YOUR top performers → copy modeled on what already converts in your account. step 5: upload ads directly to your account → new creative + copy → live in @Meta Ads Manager → no more downloading, formatting, clicking through the upload flow → agent handles the entire publish cycle step 6: content concepts + morning brief → spots patterns across winners and suggests what to test next → delivers everything to Telegram, Slack, wherever you want it → 90 seconds to read. reply "approved." done. input: your ad account + your target CPA output: an AI that monitors, kills, scales, writes, AND uploads your ads dozens of hours in ad manager → 1 text message I packaged the entire system as the Meta Ads Kit. 5 @OpenClaw skills: - meta-ads (daily checks + auto-pause) - ad-creative-monitor (fatigue detection) - budget-optimizer (efficiency scoring + shift recs) - ad-copy-generator (writes variations from your winners) - ad-upload (publishes creative directly to your account) giving it away free. comment ADS + like + follow (must follow so i can DM)

Better late than never - excited to share this one and grateful for the opportunity that @hardhatcfo and I had to support this family-owned, home service business. Got to be a part of a lot of growth (both topline and accounting & finance infrastructure buildout) that culminated in a great outcome for the family.

@farrowhardy1 How exactly are you doing this ? Just giving Claude code api access to the ST database and asking these questions via the terminal ?

Claude Code is a gamechanger when dealing with ServiceTitan data. Just pulled in less >1 min: -Avg. time for each tech to create estimate on job (fast is bad because trust isn't being built) -How does close rate change based on estimates created within 30 minutes vs. +30 mins?

Isn’t the data from the CHI / ownership works case study direct proof that it does work? I know correlation is not causation - but if explained correctly, seems like it can help? Eg if you’ve been here for X years, you’ll get 3x your annual pay in a one time bonus — that’s pretty straightforward for folks

@PEoperator Yep. The people pushing this at places like KKR have no idea what the typical employee thinks....

I totally agree with this (and have actually made the exact same “quarter per hour” comment irl). Equity awards (in whatever form) are not properly valued by most employees. I have seen this dynamic in all of the portfolio companies I’ve been a part of. The frustrating issue is the award did not drive the intended behavior change and alignment. If you are allocating equity awards, be sure the people receiving a) understand it and b) value it. One caveat- I have found that awards can be symbolically powerful- people are honored to be included. This can be useful in trying to drive behavior change or appealing to alignment, but these types of awards should typically be smaller.

@OneManLBO The combination of this approach with a growing understanding of how to use Claude code —> you will be unstoppable

I've been using AI all wrong. First week on Claude has given me a real education. For the longest time, I though it was all about prompts (explicit detail, lots of instructions, "think carefully", the "make no mistake" running joke). Nope. Or at least not exactly. I now realize it's much more about consistent context. When you work inside a single Claude project over long periods of time, every new chat starts on a rock solid context formation - project files, instructions, learnings that re-generate autonomously every night. This thing actually develops ever-persistent knowledge of your data, constraints and macro goals. It learns you. LLMs are just prediction machines that produce an optimal solution that satisfies constraints. The more constraints the thing remembers about your project, the better the answers. So you want to give this thing a sh*t ton of constraints, really. Previously, idiot me was using GPT as a point solution for every request. I felt like I had to constantly re-explain myself and what I was trying to do. To be fair, maybe I just didn't use it optimally (they have projects too). Very, very decent to high chance I'm just a dummy. With Claude, I can see the strategic vision. love the idea of this thing getting ever smarter about my past trajectory, data, goals I'm optimizing for. Feels increasingly like a strategic advisor on steroids. Very bullish on Claude as a long-term winner in this AI race, for a number of reasons. Canceled GPT Enterprise subscription last night. And man, it feels GOOD.

@toddsaunders @n8lenahan I’d assume he’s quoting gross — but I’d be curious what it cost him to build said internal tools

@n8lenahan Is that net or gross

Already cut nearly $11k in software costs per month this year and I don't think we will be stopping there. Mix of software bloat and its easier to build internal tools ourselves these days. Crazy times

Honestly no reason the sell side CIMs should even include those projections — waste of everyone’s time The irony is there are independent sponsors and searchers that are putting in projections even more aggressive than the sell side banker so their deal “pencils” to the 30%++ net IRR that those investors think they deserve

Every sell side CIM includes forecasted financials with the classic “hockey stick” projections I can only recall 2 sell side transactions I was apart of during my investment banking days where actual results outperformed the CIM projections For this reason, private equity buyers always heavily discount or ignore entirely the forecast period beyond the next twelve months

Honestly no reason the sell side CIMs should even include those projections — waste of everyone’s time The irony is there are independent sponsors and searchers that are putting in projections even more aggressive than the sell side banker so their deal “pencils” to the 30%++ net IRR that those investors think they deserve

@toddsaunders I’ve found it’s even better when you say you’re 3 years old — apparently 5 year old comprehension is equivalent to Sonnet 4.5 knowledge

my favorite thing to do in claude code

I’m either misreading your comment or Ed’s post? I read Ed’s post as saying the GTMEs are protected because they are the consultants / strategists and tools like like clay are “agents” in the sense that they’ve automated a lot of the manual work that outbound folks used to do 10+ years ago (eg clean crm records) Whereas I read your post as GTMEs will be dead in 2-3 years and clay as a software is too complicated relative to the basic motion of doing outbound which can be done more simply in many other ways

Ed knows 1000x more than I do about this stuff and my experience is this role exists bc the product is far too complicated for what it produces. Genuinely curious if this is still around in 2-3 years.

@PadraicMcC Can we please get the recording ?! I wasn’t able to make it but really want to watch it back

Got a good question on the webinar today around the risk of being a small fund and a first mover on the ai stuff (budgets are small = relatively big dollar allocations = higher risk). We thought a lot about this. My two cents: there is so much work to do in the initial phases to get the architecture / mapping / hygiene right = a lot of “easy” workflow wins that won’t go away (for the most part) as the technology advances. Sure, will our core tools evolve? Maybe. But that risk will always be present.