How I use iMessage and AI to run my life
TL;DR
Lindy is pitching itself as the ‘iMessage-native executive assistant’ for non-technical users — founder Flo says setup takes “2 minutes” with just a phone number and Google account, then it starts triaging email, prepping meetings, updating calendars, and sending proactive prompts in iMessage.
The product’s core bet is opinionated automation, not infinite flexibility — Flo repeatedly contrasts Lindy with OpenClaw and Claude/Code ecosystems as “Mac OS vs Linux/Android,” arguing that busy founders and operators want something secure and out-of-the-box, not a tool they spend weekends configuring.
The demo centered on surprisingly high-context, proactive actions — examples included triaging 63 overnight emails, drafting 4 replies, catching that Gary Danko was closed on Tuesday and suggesting nearby Quince Seafood, and Slacking a missing teammate after a meeting based on a voice instruction.
Lindy’s memory layer is the real engine — by ingesting Gmail, calendar, Slack, Google Drive, Notion, meeting transcripts, and more, it can answer questions like what a prospect does, where a team is based, or whether a law firm used the wrong billing address, making it function like a searchable ‘second brain.’
Greg mapped Lindy against a real human assistant and found strong overlap on research, scheduling, and lead follow-up — Flo’s answer was basically “yes, yes, and yes,” with built-in scheduling plus custom instructions like “after my meetings, update my CRM” and proactive meeting briefs before podcasts and calls.
Pricing starts at $49/month, but the power users are doing weird, expensive agentic things — Flo said the top 1% of users drive more than 50% of token spend, citing one finance user who has Lindy wake up every 15 minutes to hunt for dropped reservations at impossible-to-book New York restaurants.
The Breakdown
Greg frames Lindy as the ‘OpenClaw killer’ for normal people
Greg opens with a very clear thesis: lots of people admire agent systems like OpenClaw, but what they actually want is something secure, useful, and simple enough to live inside iMessage. He says he invested in Lindy and brought founder Flo on to show why this could be the executive-assistant product that regular operators actually use.
Flo’s big claim: an AI executive assistant in two minutes
Flo says Lindy exists because users kept building the same thing with Lindy’s workflow product: calendar management, meeting workflows, inbox help — basically an executive assistant. So they productized it into “Lindy Assistant,” an AI that lives in iMessage, connects to email, calendar, Notion, Google Docs, Slack, and more, and proactively surfaces opportunities to save time instead of waiting for prompts.
The iMessage demo: lower-case tone, jokes, and real work getting done
The most memorable part of the demo is how human the thread feels. Lindy sends Flo a morning brief with weather, meetings, and the note that it triaged 63 overnight emails and drafted 4 replies; it also notices his Gary Danko dinner is impossible because the restaurant is closed on Tuesdays and suggests a nearby replacement, then jokes that showing up to an empty restaurant “would have sucked.” Flo says they obsessed over that informal voice, even though model habits like em dashes are still hard to kill.
Lindy as second brain: meeting memory, Slack follow-ups, and CRM updates
Once connected, Lindy ingests the user’s digital exhaust and turns it into memory. Flo shows it recalling details from an earlier meeting with Carnegie, then demonstrates a live use case where he dictated in a meeting that Lindy should mention Ali in Slack afterward about an eval-related issue — and it did. He also shows how lightweight customization works: text Lindy “after my meetings, update my CRM,” and it asks which CRM you use and wires itself in.
Greg pressure-tests it against a real human assistant
Greg admits he already has a human assistant handling three buckets: research, scheduling, and lead follow-up for his agency Late Checkout Studio. Flo’s answer is that scheduling and email are the obvious native jobs, but research is also strong: Lindy prepped him for Greg’s podcast with notes like Greg being CEO of Late Checkout Studio and a former advisor to Reddit and TikTok. Flo also uses an iPhone action-button shortcut to send voice questions into Lindy, like asking where Henry said his team was based; Lindy searched meeting transcripts and answered: Singapore and Hong Kong, with some still in Shenzhen.
The product positioning: Mac OS for overwhelmed operators
Greg asks directly how Lindy compares with OpenClaw and the broader Claude/code-agent ecosystem. Flo is unusually candid: OpenClaw is more powerful and versatile because it can effectively modify its own capabilities, but that also makes it riskier; Lindy is less flexible by design, more secure, and easier to use. The metaphor he keeps returning to is Linux vs Mac OS, then Android vs iPhone: technical users may prefer maximum control, but Lindy is being built for the “chief everything officer” — the overloaded founder, real estate agent, sports bar owner, or roofing company operator who just wants it to work on day one.
What it can do now, where it’s going, and the weird whale behavior
Flo says the five-year vision does expand beyond executive assistants into role-specific agents, but the team is deliberately resisting that pull for now. He shares one example of Lindy hearing in a weekly business review that support was overloaded, then offering to source 100 support candidates in the Bay Area and draft outreach because it already knew the company and how to pitch it. Pricing starts at $49/month for most people, but the biggest spenders use Lindy for odd jobs like vibe coding or checking restaurant reservations every 15 minutes; voice replies exist now, phone calls are coming, and future group chats will let a human executive assistant and Lindy work side by side.