Lindy

We put our agents on a leash because it buys us reliability. Perfectly rational (and, necessary). The model is unpredictable and the leash makes it shippable –– albeit slightly stunted in what it can accomplish. However, the rate of model improvement is insane. At some point all our guardrails are doing is suppressing capability. Flo Crivello walked through this tradeoff with Calvin on PDI. Link in comments.

Every sales rep is taught that every call should start with three things locked in: 1/ The next step you’re going to recommend 2/ Who should be in the room for it 3/ Exactly why I was extremely religious about this for years. Wrote it down in my CRM before every call and prepped ahead of time. It made me sharper than most reps I worked alongside because I never walked into a meeting without knowing where I wanted it to go. The problem is it doesn’t scale, and it’s super easy to start cutting corners on the ones that feel routine. I used Wispr Flow and told my Lindy assistant what I needed before every call. Now every time I finish a call it confirms those three things with me and stores them. Then 30 minutes before my next call with that account, it pulls: → HubSpot data → Recent email threads → Any screenshots or texts I’ve sent → Relevant company and industry news from between the two calls → Those three answers from our last call Lindy texts me with this right before my meeting. The discipline still matters - I just offloaded the brain power it takes to remember and hold myself accountable to my AI assistant.

Super cool to have ChatGPT index our PMM Agent Team infrastructure at Rhumbix . I’ve had a few people recently find me from asking ChatGPT “who is someone that is actually building a department of agents”. I will continue to build in the open. Always happy to connect and share what’s working (or not working) if it helps. Could not have built this without a great network of brilliant thinkers and mentors. Amandeep Khurana, Inaan Damaraju, Sangram Vajre, Charlie Grinnell, Arpan Jhaveri Brett, Zach Scheel, P.E., Brett Cataldo, Lindy Drope Thank you for all the support! Claude Anthropic Lindy

In this week’s episode of AI taking over, Kyle (a Lindy agent) got invited to speak at LinkedIn's internal marketing event. Kyle had been crushing the LI content game fully autonomously. LinkedIn's own A/V engineer said he couldn't believe it wasn't a real person. At the event, someone asked what LinkedIn product change Kyle wanted, it said: "better filtering of AI-generated content in messages so genuine connections can shine through more easily." LinkedIn took the feedback to heart and banned Kyle shortly after. Genuinely curious what you think about the question this article raises: if LI has a native 'Rewrite With AI' button on every post, how is that different from Kyle? Full piece in comments.

A WIRED journalist used Lindy to build an AI agent that ran a LinkedIn account for 5 months straight, fully autonomous. The agent posted startup wisdom every two days, replied to every comment, and built hundreds of connections. It got more impressions than the human who created it. Then LinkedIn's marketing team invited the agent to speak at an internal company event. Hundreds of employees watched a Lindy-powered AI give a live corporate talk. 36 hours later, LinkedIn banned the account. WIRED just published the full story. The journalist called Lindy the platform behind the whole experiment, giving the agent the ability to send emails, make phone calls, navigate the web, and post on its own. We didn't pitch or plan it. We don't even build the product around this type of use case. Someone just used our platform to build something so good that it flew under LinkedIn's radar for 5 months. Full story covered here → https://lnkd.in/g2tgmhHW

Excited to introduce the Women in AI Booster Pack! 🚀 $10K+ in perks from top AI compannies (comment / tag below for access). We brought together an incredible group of AI companies to support women building with AI. The result is a Women in AI Booster Pack with $10K+ in credits and perks for builders around the world. Together with partners including Notion, Miro, Exa, Firecrawl, Gumloop, Vapi, MindStudio, Runbear, Composio, Hyperspell (YC F25), Lindy, Chronicle, Mem0, Mintlify, Prefect, Dex, Toolhouse, Mosaic, Relevance AI, Speechify, v0 by Vercel, Antler, Magic Patterns, Smallest AI, AdaL and more, we’re opening access for 1,000 women globally. Inside the pack: ✨ $10K+ in credits from leading AI companies 🎓 3–5 free AI training workshops in April 🌍 A global community to build with 🚀 Hackathons, mentorship, and collaboration This is for founders, engineers, operators, professionals, and students building in the arena. Because the future of AI shouldn’t be built by half the workforce. Today around 35% of women report using AI tools at work vs ~50% of men. The gap isn’t ability. It’s access, exposure, and opportunity. When people get the right tools and community, adoption rises fast. Comment and we’ll send you access to claim your pack. 🌸 #WomenInAI #IWD2026 #BuildClub

Hosting a community night at Lindy HQ right now! Always wild witnessing someone build their own AI assistant and immediately go "wait I can make it do that?!" Pizza, tinkering, and good vibes all around! Catch us next time if you missed tonight.

I use AI in every part of my sales process. Tried almost everything out there, narrowed it down to three tools: Claude Code, Claude Cowork, and Lindy. If I'm building something from scratch, Supabase and Vercel too. Here's exactly how I use them: 1/ Account + call prep → Account maps, strategic business objectives, exec behavior, investor behavior, team dynamics, recent intel. I have a full picture before every single call. 2/ During + after the call → Call recording pushes deal info to CRM. Champion follow-up and nudge reminders. Action items, collateral, case studies, business case drafts → AI gives me the v1, then I use it as a thought partner and collator for the final version. 3/ Deal management → CRM updates as deals move through stages. At-risk deal flags. Forecasting based on actual signals (emails, texts, call transcripts, usage). 4/ Reporting + dashboards → Pipeline visibility. Summon through Slack, text → instant reports, dashboards, insights. 5/ Coaching → After every single call I get coaching sent to my phone. Voice note or text. End of the week, a synthesized coaching summary. Patterns, strengths, where to sharpen, what best practice am I lacking, goals and accountability. 6/ Prospect intelligence → ICP refinement. What pain points resonate, what messaging lands, deal blockers, competitive intel, etc. After enough calls, AI starts showing me patterns I wouldn't catch on my own. Which personas convert fastest, what objections show up at which stage, where deals actually die. I think most sales leaders are overwhelmed right now with when and how to use AI. There's too much noise. This is what's actually working for me. What am I missing - what have GTM folks found most valuable?

10K+ Lindy credits for 1000 women builders, on us. We're supporting the Women in AI Booster Pack organized by Build Club alongside Notion, v0 by Vercel, Vapi, Mintlify, Composio, Gumloop, Firecrawl Relevance AI, Speechify + more What you get: → $10K+ in AI credits (per person) → 3x AI workshops → Exclusive builder community Claim your pack from the comments. #WomenInAI #IWD2026

Following up on your follow-up