Startups

I started my first venture at 21 and sold it in a multi-million-dollar transaction when I was 26. Today, I’m building my second venture, GreyLabs AI, where we’ve raised millions in capital and are scaling fast. Across this journey, here are 10 learnings that have stayed with me: 1. Startups fail on Day 1, not when they run out of money - if the founding team isn’t right. You need shared vision + complementary skills. 2. Sell first, build later. Don’t waste years building what no one wants. Get customer buy-in before writing code. 3. Focus on a small market. At GreyLabs AI, we’re laser-focused on the India BFSI. Shiny distractions exist everywhere, but focus wins. 4. High-impact individual contributors matter. In the AI era, such people can drive more value than larger teams. 5. Hire a strong law firm. Bad contracts (customers, vendors, shareholders, even cofounders) can destroy your business. 6. Customers > Investors. Be frugal, get paying clients, and build proof. Investors follow traction - otherwise, you’ll only get "advice". 7. Control your spending. Many founders start spending mindlessly after raising money. Waste money today, and money will waste you tomorrow. 8. Culture beats perks. My first startup was bootstrapped, with limited benefits. But no toxic managers, genuine care, and empathy earned us a 4.6 rating on Glassdoor (75+ reviews, all organic). 9. Undercommit, overdeliver. Never oversell. Promise only what you can deliver - and then exceed expectations. 10. Stay grateful. To your team, customers, investors, and family. None of this happens alone. Startups are tough, but with the right principles, they’re also the most rewarding journeys one can take. #startups #business #entrepreneurship

6 things that separate founders who scale from those who stall. I've met hundreds of founders over the last decade. Some have built extraordinary companies. Others, with equally good ideas, never got past a certain point. The difference is rarely the idea. It's almost always the founder. ➡️They hire people who intimidate them. Founders who scale are not threatened by talent. They actively seek people who know more than them in specific areas. Founders who stall, hire people they can manage easily and that's a big mistake. ➡️They kill their darlings early. A product, a campaign, a strategy they personally love — if the data says no, they let it go. Stalling founders hold on too long because their identity is tied to the idea. ➡️They stay close to the consumer even when the company gets big. At Honasa Consumer Ltd., I still read reviews. I still look at what people are saying in comments. The moment you outsource your consumer understanding entirely, you start making decisions in an echo chamber. ➡️They build systems, not just momentum. Early-stage hustle can take you to a point. But scaling requires processes that work without you in the room. If everything depends on the founder, it's not a company yet. ➡️They are comfortable with being uncomfortable. Every new stage of growth feels like starting over. New problems, new pressures, new skill sets required. Founders who scale lean into that discomfort. The others avoid it and plateau. ➡️They know the difference between being busy and making progress. A full calendar is not a strategy. The founders I admire most are ruthlessly focused on a small number of things that actually move the needle. Scaling is not about working harder. It's about thinking differently at every new stage. What would you add to this list? #FounderMindset #Entrepreneurship #Startup #D2C

9 Months of a Startup Journey: What I Learned When I Chose to Stop? After 9 months, I chose to stop. For me, it was responsibility, not giving up. Like poker, I looked at the cards, the chips in the stack, the flop, and what made sense for our team and my family. How it started We explored 5 problems, ran dozens of validation calls, warmed up VCs, and did competitive research at RSA San Francisco. We realized we were late to the market in our chosen direction; a pivot was required. 8 lessons about fundraising 1. Geography - Founders in different countries make VCs hesitate; it raises GTM and ops concerns. 👉 Takeaway: Pick a clear home base for fundraising. 2. Scale - Real pain is not enough; VCs want credible scale paths even early. 👉Takeaway: Show sharp differentiation and how you rise above the noise. 3. Israeli cyber thesis - Many Israeli cyber VCs prefer narrow plays that fit a thesis and can exit quickly; few will underwrite a new category. 👉Takeaway: Start focused within a trending space; expand later so risk vs. opportunity is legible. 4. US VCs want proof - Vision lands, but checks rarely land on a deck unless they know you well or you are a second-time founder. 👉Takeaway: Bring a working prototype and early reactions or arrive with a strong Israeli lead VC. 5. Singapore VCs at Seed - Cautious on early cyber and prefer to follow a lead once there is a prototype and first customers. 👉Takeaway: Prototype plus 2–3 design partners plus a lead VC unlocks doors. 6. Response time - Some reply the same day; others vanish. 👉Takeaway: Send a crisp summary post-meeting, nudge after 7–10 days, then move on. No pace means no intent. 7. Reading feedback - “Too early” can mean “no clear edge.” “Great team, come back with customers” can mean “we are not convinced you can sell.” 👉Takeaway: Do not take it personally; mine the tactical notes on gaps and next steps. 8. Familiar bias in Israel - Military-tech background often gets priority unless you have just exited your startup 👉Takeaway: Come with proof in the market, trusted references, and a prototype in customers' hands. 🔻 Turning point after Black Hat Las Vegas I asked the hard co-founder questions: Does the vision still align? How big are the gaps? Which moves gain chips instead of burning them? We saw it differently. The conclusion: Stop now. Stopping at the right time is a management muscle. The journey didn’t end as we hoped, but every day taught me something new, and I’m at peace with the decision. Grateful to my co-founder, to customers who believed, VCs who shared time and feedback, friends and mentors, and my family. Special thanks to my wife for her ongoing support in every risk we take. Until the next one! Full story with examples and playbooks: https://lnkd.in/g4jGs-st #startup #founders #cybersecurity #fundraising #rsa #blackhat #poker

AT&T cut their AI costs by 90%. Not by negotiating better contracts. By rethinking the architecture entirely. They were processing 8 billion tokens a day through large, general-purpose models. Expensive. Slow. Hard to scale. So their CDO Andy Markus rebuilt the entire orchestration layer around small, purpose-built models, each trained for a specific task, coordinated by a larger model above them. The result? They now process 27 billion tokens a day. At roughly 10% of the previous cost. And accuracy? Markus says small language models are "just about as accurate, if not as accurate, as a large language model on a given domain area." His conclusion: "I believe the future of agentic AI is many, many, many small language models." Most business leaders are still debating which big model to buy. The smarter question is: what specific tasks do I actually need AI to do, and can a smaller, cheaper, focused model do it just as well? In most cases, the answer is yes. Interview with Andy Markus here: https://lnkd.in/e9A6xt6Y ___ I write about AI, business strategy, and leadership for decision-makers. Enjoyed this post? Like 👍, comment 💭, or re-post ♻️ to share with others. Hit the 🔔 on my profile to receive my latests posts.

Welcome back to the era of Agile Funding 🚀 For a long time, the standard startup playbook for founders was to build a deck, clear your calendar for 3 months and take 50+ meetings to find a lead investor for a fixed size round. But recently, I've noticed a massive return to "High Resolution Fundraising" - a concept Paul Graham from Y Combinator popularized more than 15 years ago. The idea is simple... instead of raising a big round all at once, founders raise capital as they go and as they need it. This strategy was the darling of the pandemic era. Founders were stacking SAFEs like pancakes, closing checks at 10am and raising the valuation cap by 2pm. But when the market corrected in 2022, the industry hit a collective pause on this practice. Investors demanded more structure, lead terms and the 6 month fundraising roadshow came back with a vengeance. 💥 But in 2026 the pendulum has swung again and Agile Funding is back in full swing. In an AI-driven world, a 3 month diligence process is a lifetime. A 2 person team can build a feature, sign a pilot and double their value in the time it takes a traditional VC to schedule a second partner meeting. Small SAFEs allow a founder to capture that value in real time. At Everywhere Ventures we are seeing early stage founders raise $250k to hit a specific engineering milestone, then immediately open another $500k tranche at a higher cap once they’ve de-risked the tech. They aren't raising for 24 months of runway, they are raising for the next three months of velocity. My advice for founders is to consider this method but to go in eyes wide open. Many VCs hate this practice with a passion and I'd agree that stacking SAFEs can lead to a debt sandwich down the road. Plus raising the valuation cap day to day can feel overly opportunistic and rub some investors the wrong way. I totally get that. But I also believe that the rules have shifted once again with AI and opening up a steady flow of cash can be useful for early stage founders. If you have the data, the customers and the velocity, you may not have to wait for a lead investor to give you permission to grow. You can use small SAFEs to lock in the believers early, reward them with a lower cap and keep your focus on the product / customers. It's been interesting to see this method come back into fashion, let's see how it plays out this time around 👀

I recently spoke with an early-stage AI app founder who was desperate to hire sales reps because he dreaded founder-led sales. This is one of the most common failure modes I see with technical founders—and it significantly impedes the path to product-market fit. Here's how to think about the right order of operations in early sales motions: Phase 1: Prototype & Validation In the earliest stage, the feedback loop between customer conversations and product roadmap must be extraordinarily tight—making founder-led sales absolutely non-negotiable. This phase is critical because you're identifying your true ideal customer profile (ICP) and learning how to effectively communicate your product story and address common objections. As you accumulate hundreds of demo repetitions (while refining your product based on feedback), you gradually assemble a winning process. Phase 2: Founder-led Sales Scale-Up Your mission here is to create the sales playbook that will guide future reps. You need sufficient pattern recognition to understand which messages resonate with which personas. I recall meeting Desmond Lim, CEO of Workstream, several years ago (not an Emergence portfolio company, but I deeply admire what they've built). He showed me the remarkable 60-page playbook he crafted documenting their entire sales process—before hiring a single AE. Every nuance. Every objection. Everything a new rep would need to succeed. While perhaps extreme, this perfectly illustrates the principle: scaling go-to-market requires mastering your ideal sales motion before delegating it. Phase 3: Hiring Initial Sales Reps Most founders default to sequential hiring—start with one rep, evaluate results, then proceed. However, we recommend hiring 2-3 sales reps with diverse backgrounds simultaneously, enabling you to effectively A/B test different profiles. Regardless of approach, ensure these early hires are "renaissance reps" with rapid iteration capabilities rather than purely "coin-operated" sellers. Mark Leslie has a great foundational article on the Sales Learning Curve provides excellent guidance. I'll link it below. So embrace the early sales work, even when it feels uncomfortable. It's fundamental to building a foundation for lasting success.

As AI efficiency, privacy, and cost become central to deployment decisions, Indian startups are increasingly adopting Small Language Models (SLMs) instead of Large Language Models (LLMs), Vaibhavi Khanwalkar reports for The Economic Times. Startups across fintech, healthtech, and legaltech are shifting toward compact AI systems to address high cloud expenses, patchy internet infrastructure, and stricter data privacy requirements in India, the report says. For instance, wealthtech firm Stockgro and trading platform Dhan are using compact models for market analysis and financial intelligence, while legaltech startup August deploys on-premise AI to safeguard client data. Healthcare company Qure.ai runs lightweight models directly on diagnostic devices for offline clinical analysis, and construction tech firm Powerplay is using homegrown models to improve accuracy in project workflows. Gnani AI, Shunya Labs, and Adya AI are all developing smaller, real-time voice and enterprise AI systems that reduce computing costs while supporting localisation and data sovereignty needs, the report says further. Unlike LLMs that rely on vast cloud-hosted datasets processed through offshore GPU data centres, SLMs operate on smaller, specialised datasets and can run locally, enabling stronger control over sensitive information, the report adds. While global LLMs perform well on broad use cases, smaller models deliver greater accuracy for domain-specific applications, note industry experts. “Most LLMs cannot do deep fundamental or technical analysis with key market signals, so we decided to build an SLM. What’s more, it costs less and has fewer hallucinations”, said Ajay Lakhotia, Founder of Stockgro. Meanwhile, India’s push to develop sovereign AI models is gaining early validation, suggests another report by businessline. Healthcare and education institutions are emerging as key adopters of locally tailored AI solutions under the India AI Mission, even as enterprise uptake remains nascent, the report says. Companies including Tech Mahindra and Fractal Analytics have already reported strong interest from domestic and overseas institutions seeking linguistically and culturally contextualised AI applications, with use cases ranging from education-focused LLMs to healthcare chatbots and diagnostic support tools, the report adds further. What does this rising interest in compact AI models suggest for India's tech landscape? Share your thoughts in the comments section. ✍: Nakul Ghai 📷: Getty Images Source: The Economic Times https://lnkd.in/dPgjfkZE businessline: https://lnkd.in/dzCx4V4M #AI #AImodels #Startups #Technology

CEOs: You Don’t Need More. You Need More Leverage. Most scaling founders keep chasing: 🔹 More hires 🔹 More funding 🔹 More tools 🔹 More features But here’s the uncomfortable truth: - You’re not running out of resources. - You’re not using leverage. The Real Bottleneck? Misused Potential. Early-stage and scaling companies aren’t broken because they’re under-resourced. They’re stuck because they haven’t built systems that multiply effort. The best CEOs I’ve worked with don’t ask: ❌ “What do we still need?” They ask: ✅ “What’s the highest-leverage move we haven’t made yet?” That shift changes everything. Here’s how elite operators think in leverage, not labor: 1. The 80/20 Principle → 80% of outcomes come from 20% of inputs. - Cut what doesn’t move the needle. Double down on what does. 2. Strategic Delegation → Don’t delegate tasks. - Delegate decisions—and give people the context to run. 3. Compounding Moves → Invest in what grows over time: - Systems.  - Talent.  - Brand trust. - Stop resetting.  - Start reinforcing. 4. Flywheel Thinking → Build momentum that feeds itself. - One smart move should fuel the next. Quick CEO Self-Check: Block 10 minutes this week and ask yourself: - Where am I still the bottleneck? - What have we started that won’t compound? - Where are we applying effort where we should be applying leverage? Final Thought: → You don’t scale by working harder. → You scale by building systems that do the work for you. CEOs: What’s one leverage point you’ve doubled down on this year? Let’s swap strategies in the comments. ♻️ Share this with your network if it resonates. ☝️ And follow Stuart Andrews for more insights like this.

Underrated business model: Turning a service into a product. It helps solve the issue of scaling for: • Agencies • Freelancers • Solopreneurs Here’s how a "productized service" works: First, let's talk about the trap most consultants & contractors fall into: Trading hours for dollars. Even at high hourly rates, you're still capped by time. Earn more = work more. Stop working, and the money stops too. But there's a better way... Enter the "Netflix of Services" model: Instead of hourly billing, you package your expertise into a subscription. Clients pay a fixed monthly fee for access to your skills. Just like Netflix doesn't charge per movie, you don't charge per task. 7 quick examples: 1. DesignJoy • The service: Design work for $4,995/mo • The product: Unlimited design requests The keyword is “unlimited.” Whether a client makes 5 requests a month or 50, the price stays the same. So, what's the catch? Clients can submit only 1 request at a time. Each one gets placed in a queue. This natural throttle prevents overwhelm while maintaining the "unlimited" promise. 2. WP Curve • The service: WordPress maintenance • The product: Live access to a developer 24 hours a day for $59 The model worked so well, they scaled to an exit to GoDaddy in 2016. 3. ViralCuts • The service: Short-form video editing • The product: A trained editor to embed on your team (I'm part owner on this one) 4. Ninjas • The service: Accounting & tax help for ecomm businesses • Product: Fixed scope, flat monthly fee The founders originally prepared custom proposals & struggled with revenue. 10 months after productization, they reported hitting $10k MRR. 5. Contentfly • The service: Content creation • Product: A set amount of words/month → So... How can you replicate this? 3 things you need: • A skill/service. • SPEED in performing that skill. • A niche searvice offer within that skill to set yourself apart. Once you’ve picked your service: • Package it into a monthly subscription • Set clear boundaries and expectations • Use tools to automate delivery • Focus on speed of execution Just remember... The real power isn't in the "unlimited" promise. It's in the constraints you build around it: • Defined scope of work • Clear turnaround times • One request at a time These boundaries make the unlimited model actually sustainable. When you're not trading hours for dollars, something magical happens: You think bigger. And the best part? Productizing your service creates a systematized business with recurring revenue. Resulting in a real business you can actually sell someday. ↓↓↓ Think about creative ways to make money is one of my favorite things to do. Grab these resource to steal my top 13: https://lnkd.in/gkMtSr-3

There's been a massive shift in startup fundraising since 2021 - and it's been hiding in plain sight. In the before times (pre-2020), startups raised mostly primary rounds. And when they did raise bridge funding, it was usually through a priced equity financing (with a valuation, price per share, all the normal stuff). Those days are gone — today, most startups (especially at Seed / Series A / Series B) will raise the their bridge capital on convertible notes or SAFEs. Look at the percentages in the bars below. The dark grey is priced bridge rounds, the bright orange is convertible note/SAFE bridge rounds. Every year, the convertibles eat more and more share for bridges. So what does this mean? • Bridge round data is notoriously tricky to find and analyze, but convertible bridge rounds are even harder to lock down. More convertibles = more bridge rounds unaccounted for.    • Less valuation certainty. If a startup raised Series A at $50M post-money, but then raised a convertible note bridge round at a $60M valuation cap...what's the right valuation? Big debates among GPs, founders, and LPs on this topic.    • More cap table complexity. Convertibles (especially SAFEs) can feel "easier" to raise since they are one-off and require less legal time...but that's just upfront. Upon conversion in the next round, they can inspire headaches.    • Tale of two cities. Most bridges are done when the company is not doing so hot. They are lifelines towards the hope of another round or potentially an exit. But some are done because the company is so damn hot the initial investors want back in quickly (the "pre-emptive" bridge). Which is which? Pros and cons to this change, but the shift is undeniable. More and more venture dollars flow through convertibles of all kinds, avoiding the traditional priced round structure. #startups #founders #bridges