Dariel Vila

The “big beautiful bill” will impact the software engineering job market in a positive way especially for startups! Between 2022 and 2025, startups had to amortize the cost of their software engineering talent. Here’s an example: Say your business made $100,000 in revenue in 2021 and you paid a software engineer $100,000 to maintain your website. In 2021, you would be showing no profit and owe no tax. In 2022, this changed. That same company had to “amortize” the cost of software engineering as research and development costs over 5 years. So in year 1, you would have $100,000 revenue and $100,000 in cost. But you can only deduct $20,000 from your taxes. So you’re taxed on $80,000 and probably owe the IRS ~$20,000 with no money to pay them! Many companies in 2022-2024 had to take out loans to pay their tax bills because of this! We’re going back to how it was prior to 2021 with the exception of offshoring talent. Offshored R&D must still be amortized. This change should bring back many software engineering jobs to America and reignite the job market. There is hope that this dog water job market will improve!

689

95 Comments

SaaS monetizes features. Rails monetize mistakes. The main objection to Rails: "Isn't this just network effects?" No. Rails don't create network effects. They create learning effects. Every time work flows through the rail, it doesn't only deliver an outcome, it makes the rail smarter. SaaS scaled revenue. Consulting scaled judgment. Rails scale error correction. While LegacyCos bolt AI agents onto existing workflows, NewCos build differently. They know you can't buy five years of resolved edge cases. That compound learning becomes the moat. Learning Effects > Network Effects Classic SaaS defensibility came from network effects. Slack gains utility with more teammates. But SaaS compounding is bounded. Each instance learns locally. Operating Rails compound differently. Every GitHub Copilot rejection, every Stripe fraud flag, every CrowdStrike attack blocked makes the entire network smarter. Expensify example: When you correct a miscategorized expense, it only improves your workspace, despite 15 million users generating errors daily. Imagine Expensify as a Rail: every correction across thousands of companies teaching the entire network. Your operating costs benchmarked against similar companies. Edge cases from one company preventing errors at another. This isn't about user count. It's about work count. Every pattern resolved becomes reusable logic with provenance. Rails monetize mistakes across the network. SaaS monetizes features in isolation. The Three Laws of Operating Rails Law 1: Learn faster than they copy  Your rail must improve faster than competitors can imitate. A competitor can copy features in weeks. They can't compress five years of edge cases resolved, versioned, and rolled back. The moat isn't what you built; it's how fast you compound. Law 2: Make it tacit, not portable  Perfect documentation is easily copyable. The real moat lives in tacit knowledge, patterns that only work with your specific context, governance, and audit trails. Law 3: Power without transparency kills trust  As rails automate more decisions, they need more governance. One bad auto-execution can destroy years of trust. Constitutions, vetoes, one-click rollbacks aren't nice-to-haves. They're essential. NewCo Playbook: Start Boring, Compound Fast (Exclusive to newsletter subscribers) The winners won't be the ones who automate fastest. They'll be the ones who learn from failure fastest. LegacyCos are adding AI agents to workflows, hoping for magic. But intelligence in local instances doesn't compound. You don't build moats by making each silo smarter. NewCos who grasp Operating Rail laws will target boring shared burdens. They'll turn every error into network intelligence. They'll shoulder bounded liability to earn trust. And they'll compound learning at rates LegacyCo can't match. Strong operators compound errors into moats. Weak operators add features and hope. (Full version sent to subscribers)

55

24 Comments

We connect brands with engineers (5K LinkedIn followers) Then outperform agencies with $1M retainers Here's the framework: 1: Find engineers posting in their industry 2: Structure $2-5K/month partnerships 3: Get 8x engagement vs company pages 4: Reduce CAC by 30-40% with authentic reach Recent Wins: AI Cybersecurity startup → Partnered with: Security engineer (6.5K followers) → Added: Weekly vulnerability breakdowns → Result: 23 qualified leads vs 8 from paid ads Developer tool company → Partnered with: vibe coding expert (3.8K followers) → Added: "Building in production" content → Result: $38K pipeline from 3 posts Data platform → Partnered with: ML engineer (8K followers) → Added: 15 migration use cases by industry → Result: 11 enterprise demos vs 2 from Google Ads Why this works: Micro-experts beat generic macro-influencers in B2B. Buyers want: → Peer perspectives, not vendor content → Technical depth, not surface features → Problem acknowledgment, not just solutions → Practitioners, not evangelists You're not competing on reach. You're competing on relevance. The Synnc advantage: Match brands with domain experts. Not generic influencers. Engineers who solve these problems daily. That's why it works. The Reality Check: Every B2B brand can tap 20+ domain experts tomorrow. → Engineers documenting their stack → Architects sharing decision frameworks → Developers building in public The distribution moat is there. Most just can't see it. Look into Synnc

713

220 Comments

💡 This RAG Agent in n8n is built to handle real knowledge, not just chat. It uses GPT-4o for responses, Supabase for vector search, Postgres for memory, and Google Drive to auto-load new docs. Ask it anything - it pulls from your actual data and replies with context. No need to code. Just connect, upload, and start chatting with your knowledge base.

220

163 Comments

Founders NEED to be smarter about structuring POCs. A pilot is not “prove the tech.” It’s 30 days to co-create a business case (ROI model) with your buyer. Do this: 1. Charge for the pilot. It filters tire-kickers. Be explicit: pilot fees ≠ production pricing. 2. If pushed on price before value: -Anchor to ROI: “For peers we unlock ≈ $10M; we target 1:10 ROI.” -Give ranges, not a point: “Final pricing will land between $500k–$1M depending on realized value.” 3. Build the ROI model with the buyer (during the pilot): Incremental gains: revenue lift, churn reduction, throughput, win-rates. Cost savings: headcount/time saved, license/tool consolidation. Opportunity cost: what higher-value work replaces the saved time? The Outcome? You qualify faster, negotiate on value (not features), and set up a clean path to outcome-based pricing. Save this for your next pilot. #AI #SaaS #Pricing #Monetization #GTM #VerticalAI #Founders #Sales

302

28 Comments

MCP gave agents memory. A2A let agents talk to each other. Now comes AG-UI — the protocol that brings agents to life on your screen. Most agents today live in the shadows. They migrate data. Fill forms. Summarize docs. And vanish. But the next generation won’t be invisible. They’ll be interactive. Visual. Present. Sitting inside your app—reading the room, adapting to you. Enter AG-UI: Agent-User Interaction Protocol An open spec for bridging: 🧠 AI agents (LangGraph, CrewAI, OpenAI, etc.) User-facing apps (React, Next, Electron—your stack) It’s a two-way pipe. Simple, open, and built for real-time collaboration. To make this possible, agents need to: • Stream updates in real time • Orchestrate tools with feedback • Sync with the frontend’s state • Respect app-level security walls • React instantly to user actions AG-UI makes all of this possible— using event-streaming, not complex APIs or polling. The flow (in 5 steps): 1️⃣ Your app sends a user request 2️⃣ It opens a persistent event stream 3️⃣ The agent emits work-in-progress updates 4️⃣ Frontend receives and renders each update in real time 5️⃣ The app and agent stay in sync—like a live teammate, not a black box This flips the agent model: From back-end bots… To on-screen copilots. And it works across any stack: SSE, Webhooks, HTTP — whatever your infra prefers. We’ve had agents behind the scenes. AG-UI brings them on stage. Follow Gaurav Bhattacharya for more on Agents Post Credit - Alex Wang, Image credit: JoinDailyDoseofDS, Github repo in comments

341

106 Comments

$10M ARR is the FU MONEY of SaaS. Here’s the breakdown: I think $10m ARR bootstrapped is the new unicorn. Why? Because if you can bootstrap $10m ARR in a lean way, all your wildest dreams will come true. You will be able to do whatever you want with your life. (I know because I’ve done it) Not only that, I think bootstrapping $10m ARR is actually way easier than creating a unicorn. And WAY less risky, with a FAR higher probability-weighted financial (and happiness) outcome. Here’s 6 reasons why all founders should commit to bootstrapping $10m ARR: 1. You and your co-founder will be making over $1m/yr in salary and dividends. This is simple math (another benefit of a small, efficient team). You and your co-founder can easily clear $1M each at $10M ARR, and probably more. With VC money, you have to be much, MUCH larger before you can take home that kind of money. It will be years later, if it ever happens. 2. You can sell your business in ONE SECOND for $30-$40m (or more) I’m sure you’ve all seen the math. Dilute yourself down to 5% on a 15 year run to $1B, and if you didn’t pull out HUGE secondary checks, you actually come out behind. Not only that - there are only a handful of buyers at $1b. There are hundreds - maybe thousands - of EBITDA buyers at $30-$40m. 3. You can hire a CEO to run your company, work 1 hour per week, and clip the coupon 4. You can work 100 hours per week and run yourself into the ground (if that’s your thing) 5. You can go raise $50m and build a unicorn Your product and GTM will be so tight that if you want to hit a home run at this point, you can absolutely do it. And as a profitable company, you will have leverage to get good terms from growth equity funds. Many will put the money straight in your pocket. 6. You can keep running it lean, get to $25m ARR with 25 people and pay yourself $10-$15m per year. This is a possibility that literally no one talks about. And again, I think it’s actually much easier and higher probability than creating a VC-backed unicorn. Exit or not, if you get here, game over. You get a decent exit every year! TAKEAWAY: If you can bootstrap $10m ARR, the options available to you are SO GOOD. The harsh reality is that most companies never get there. Why? Because you get traction and VCs show up at your door. They gas you up with hopes and dreams about how your SaaS will be a decacorn They give you money that you don’t need, which you promptly turn around and spend. All the while, it’s actually INTERFERING with the one thing that will make you great, capital or not. Product-market fit. If you can find PMF, bootstrapped, then use money from your customers to get to $10m… You can do anything you want with your life. Ross Andrew Paquette, who grew Maropost $0-$25M ARR profitably in 3 years, is the PERFECT example of this. He's joining me tomorrow at 3pm ET on UnF*ck My Startup LIVE to share how he did it and his growth playbook for 2025. Join us: https://lnkd.in/gAe5qsBy

814

121 Comments

Your auth, your rules. Here’s how you can do this with FusionAuth. Don't let multi-tenant SaaS auth providers hijack your roadmap. They force you to fit into their environment and timeline. FusionAuth is different - it's a developer-first CIAM platform you run on your own infrastructure. 1. 𝗗𝗼𝘄𝗻𝗹𝗼𝗮𝗱𝗮𝗯𝗹𝗲 𝗖𝗜𝗔𝗠: Run FusionAuth in your environment with full auth features (login, registration, SSO, MFA, etc.). 2. 𝗦𝗶𝗻𝗴𝗹𝗲-𝘁𝗲𝗻𝗮𝗻𝘁 𝗵𝘆𝗯𝗿𝗶𝗱 𝗱𝗲𝗽𝗹𝗼𝘆𝗺𝗲𝗻𝘁: Run anywhere (local dev, CI/CD, any cloud or on-prem) with dedicated resources that scale with you. No noisy neighbors. 3. 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗲𝗿-𝗳𝗶𝗿𝘀𝘁 𝗔𝗣𝗜: Every feature is accessible via API for easy integration. 4. 𝗡𝗼 𝗺𝗼𝗰𝗸𝘀 𝗼𝗿 𝗰𝗼𝗻𝘁𝗲𝘅𝘁 𝘀𝘄𝗶𝘁𝗰𝗵𝗶𝗻𝗴: Use the real auth service in dev/test, just like in production. 5. 𝗨𝗽𝗱𝗮𝘁𝗲 𝗼𝗻 𝘆𝗼𝘂𝗿 𝘁𝗲𝗿𝗺𝘀: You decide when to upgrade, not the provider. No surprise downtime or breaking changes. Get started with FusionAuth: https://fnf.dev/4jLmt02 Huge thanks to FusionAuth for sponsoring this post.

254

28 Comments

LLMs and Codegen have collapsed the time and cost of building new software and much can be generated. Now, more emphasis is needed on the ability to scope requirements and write specifications. Specifying intent can be incredibly difficult given the fuzziness and ambiguity of natural language. Maybe the new 'time to build' is actually 'time to think'.

30

6 Comments

5 'no-brainer' tools YC alums use to fill calendars on autopilot. Playbook: Build a machine to find AND get found by customers. ____ 1. Outbound Stack "How are you finding customers?" Clay - find leads with enriched data HubSpot - send emails and track deal pipeline HeyReach.io - linkedin and multi-channel outreach ____ 2. Inbound Stack "How are customers finding you?" Relixir - fastest way to show up on ChatGPT when prospects search Framer - ship landing page and host blogs that rank for search engines Any tools I'm missing? Comment below.

79

9 Comments

Starting today, our engineers stop writing code. For the next 30 days, we're moving our entire team — engineering, GTM, and design — from writing and reviewing code to providing architecture and scaffolding for agents to take features from spec to production. We're going to fail. Probably a lot. But we're doing it anyway. Here's why: - A few days ago at a True Ventures founder gathering, I watched teams demo what AI-first SDLC actually looks like now. Not in theory. In production. - 3 engineers doing 2,000 commits/month. - GTM teams of ONE producing thousands of pieces of content, measuring impact, and iterating — all using autonomous agents. This wasn't possible even 3 months ago. The advances in coding agents have crossed a threshold that changes what a small team can do. And if you're not experimenting with this right now, the gap is growing fast. We're going to share what works and what doesn't along the way. If you want to follow along, we'd love the company. #aifirst #ainative

195

40 Comments

BREAKING: Google just dropped MindMaps for NotebookLM. It instantly turns your notes into visual diagrams Here's how it works: Go to NotebookLM, add sources, click ‘Mind Map’, and done! • Transform chaotic ideas into visual maps • Get AI-powered suggestions • Collaborate with team members in real-time • Integrate directly with your existing Google Workspace ecosystem

188

6 Comments

Many SaaS companies are already extinct. They just don't know it yet. I've watched AI compress decision cycles from months to minutes. What some see as a tooling shift, I see as a systems failure. Because…Most SaaS was built to capture decisions not make them This take competition and take it to a whole new level... to extinction-level speed. I used to think 10x better would win, now I'm seeing 1000x faster eliminate entire categories. When we started this company, we optimized for quality. Now speed is the product - - and quality is table stakes. The companies that get this are building the future, the ones that don't are becoming ghosts. The winners understand one thing: how quickly your system of record can turn insight into action. Here are the 6 mindsets that determine who survives: 𝗖𝗼𝗺𝗽𝗿𝗲𝘀𝘀𝗶𝗼𝗻 𝗧𝗵𝗶𝗻𝗸𝗶𝗻𝗴 Stop building for quarterly reviews. Start shipping in real-time cycles. → Ship a feature update today, not next sprint. 𝗦𝗲𝗰𝗼𝗻𝗱-𝗢𝗿𝗱𝗲𝗿 𝗧𝗵𝗶𝗻𝗸𝗶𝗻𝗴 Stop solving today's problems. Start preventing tomorrow's friction. → Kill the workflow before adding another integration. 𝗦𝘆𝗻𝘁𝗵𝗲𝘀𝗶𝘀 Stop sending reports about data. Start shipping decisions from insights. → Replace your dashboard with automated actions. If a human has to notice it, it's already too late. 𝗦𝘁����𝗿𝘆𝘁𝗲𝗹𝗹𝗶𝗻𝗴 Stop explaining what you built. Start showing what customers achieved. → Share customer wins, not feature releases. 𝗗𝗲𝗰𝗶𝘀𝗶𝗼𝗻 𝗠𝗮𝗸𝗶𝗻𝗴 Stop optimizing for consensus. Start optimizing for velocity. → Skip the committee. 𝗣𝗿𝗼𝗯𝗹𝗲𝗺 𝗦𝗼𝗹𝘃𝗶𝗻𝗴 Stop adding features to fix symptoms. Start removing the root cause entirely. → Delete the form instead of improving it. Most SaaS was built to observe, but now it needs to act. In a world where decisions happen in real time, systems that only record reality won’t survive it. Speed isn’t an advantage anymore. It’s the baseline for existence.

67

24 Comments

Founder mode isn’t enough. Vercel’s founder Guillermo Rauch believes it “fundamentally doesn’t scale.” Instead, he practices what he calls “recursive founder mode.” In Every Inc.’s podcast, he explains that recursive founder mode works on 2 parallel tracks to increase “founder energy” in a company: (1) Attracting founder-level talent (the high-ownership people who’d probably go on to start companies themselves) (2) Creating an environment that nurtures talent, where ambitious builders can thrive Definitely easier said than done, but Guillermo Rauch believes his job as a founder is to scale other founders, embedding the instinct to move fast and push boundaries into Vercel’s DNA. To create a truly scalable business, you first need to create a strong talent org. We took a deep dive into Vercel’s talent strategy, including: 1/ How they built a company that operates the same way their product behaves 2/ The innovative way Vercel structures roles within their org 3/ How their employer communicates their product philosophy without a single ”who we are” bullet point 4/ Concrete tactics to implement from Vercel’s talent playbook When you’re pursuing explosive growth while trying to fundamentally change how people build and experience software, maybe regular founder mode isn’t enough. Full story linked in comments.

80

3 Comments

By the end of the YC W24 batch, Arman and I were stuck. We’d been building a tool for construction teams that no one really needed. We’d been so heads-down on the product that we missed the bigger picture: we didn’t know the space or the real problems that existed. It became painfully obvious that it wasn’t going to work. So we made a call: whatever we did next, we’d go deep. Talk to everyone we could, understand what they actually care about, and experience their problems first hand. We had to become insiders. We also knew that if we were going to spend the next decade building something, it needed to be in an industry we genuinely cared about. Growing up in Silicon Valley meant tech was always around us, but our time at UIUC—surrounded by cornfields and factories—gave us a unique appreciation for industries quietly powering everyday life. I thought back to my time at Palantir, where I built AI for a Fortune 50 manufacturer, and saw how transformative LLMs were going to be for industrial companies—but also how the space was being overlooked by my peers in tech. So, we started by talking to the people doing the real work. First week, we went door to door to 27 local building suppliers in SF. We got hooked. In the next 6 months, we visited factories, warehouses, and offices across 13 states and met with 263 C-suite execs. We found ourselves in rooms with everyone from Fortune 100 executives to family-owned mom and pop shops.  Here are 5 lessons we learned from 263 conversations: 1️⃣ People believe your story, not your product We spent a long time experimenting with tons of different types of outreach and sales pitches. But the meetings where we just laid out why we were there and what we were chasing went the furthest. 2️⃣ Listening > Talking The real wins came when we just let people ramble and go off on tangents. That’s when we learned about the problems worth tackling, or the ones they didn’t even care about. 3️⃣ Ask the right questions We started out asking questions like “What’s your biggest problem?” This didn’t work. It was too broad and most people just spit out something generic. When we started asking things like “What’d you do last week when that order got delayed?” they couldn’t fake it. We got the raw, unfiltered truth. 4️⃣ Meet in person We did everything to meet folks in person even if it meant long drives and multiple flights for a 30 minute time slot. Nothing beats sitting across from someone and having an honest conversation. Those in-person chats build real trust that we couldn't have built online. 5️⃣  The real stuff happens on the floor, not the boardroom We thought the execs would have all the answers. Nope. The best insights always came from walking warehouses or watching workers do their thing. Execs see the big picture; the floor reveals the day-to-day headaches they miss. This is just the start. Follow me for more! If you’re a manufacturer or distributor curious about AI, hit me up. Let’s talk!

195

19 Comments

"Everybody thinks it gets easier when you’re at $100M+ ARR, cash-flow positive, and growing. But I just had the worst two months of my entire life.” That’s what Marcelo Lombardo, founder of Omie, told me on Friday. Omie is the Brazilian SaaS giant founded in 2013. They just announced a $160M Series D. With 180K SMB clients and 30-40% annual growth, they’re the #1 cloud ERP and fintech for small businesses in Brazil. They are real OGs in the BR ecosystem. I picked Marcelo's brain on his learnings going from $0 to $100M ARR. "It doesn't get easier as you grow, it just gets different". I burst into nervous laugh but also in relief. The conversation was so fun and full of golden nuggets that deserve to be shared.. so here are some of the lessons and stories he shared: 1. Borrow credibility to open doors In the early days, he picked up the phone relentlessly to speak with his first customers and ask a simple question: “Do you remember where you found us?” Many said “Google", until he pressed. The real answer: “My accountant told me to use Omie.” That’s when they realized accountants, not ads, were the real channel. They flipped strategy into partnerships with them. Today they drive 80% of Omie’s sales. 2. Commissions can erode trust “If my client finds out I’m making money from this, it hurts my credibility. Give them the discount instead.” the accountants would tell him. The best accountants valued trust over margin. Incentives that work in SaaS backfire when partners are fiduciaries. 3. Trust lowers walls faster than incentives When Marcelo asked accountants if he could sell to their clients directly, they handed him their entire client list. Why? “Because they trusted me.” he said. Alignment matters more than commissions. 4. Trust scales through presence “The job to be done is not to scale software sales, but to scale trust.” Omie’s AEs visit 3-4 accountants a day, every 15 days. Not to pitch, but to listen: “What’s your problem today? What’s your client’s pain?” That cadence compounds relationships and yields warm intros. 5. Trust is long-term leverage Instead of outspending competitors on ads, Omie built a distribution moat by being the most trusted partner to accountants. Other nuggets from our call: - Trust > commissions - Give partners the product for free so they become power users - Keep partnerships alive with cadence: 3–4 visits/day - Great partners protect client trust; weak ones chase margin - At every stage, the problems change, but trust is always the multiplier If you're building a compay, or operating, I encourage you to share what you learn, you never know who you help. Shoot me a comment or a dm. Now, back to building 💪. #startups

70

7 Comments

Moltbook is redefining what AI in businesses actually looks like. 🫨 It may turn out to be just another disruptive trend, but Moltbook does something important: it shows what the agent-to-agent (A2A) model looks like, at scale. For teams building with agents, here’s what it clarifies: 1. A2A is a coordination problem, not an intelligence one The real shift isn’t smarter individual agents. It is the systems where agents learn from each other, hand work off cleanly, and build on shared context, without humans stitching every step together. 2. Autonomy without governance breaks systems Agents with autonomy means power with scope. Unsupervised, agent behaviour can be chaotic. systems must use permissions, telemetry, human approval gates for critical actions (hopefully to contain unpleasant surprises) 3. Humans move from operators to system designers As agents take on execution, the human role shifts away from task management toward setting intent, constraints, and outcomes. (This may or may not trigger the reverse flynn effect 🫣) KEEP UP!!! you snooze, you lose 😪

33

6 Comments