loading...
Day.ai rebuilds CRM from scratch around AI – no manual data entry, no schema design, just relationship context that emerges automatically.
ENTRY ANGLES
Rebuild existing product categories with AI-native architecture instead of grafting AI onto legacy structures · Offer qualitatively simpler user experience through AI-native design · Provide clear migration path from incumbent products to new AI-native alternative
VERTICALS
CAPABILITIES
AI-native product architecture design, Understanding of incumbent product foundations to identify what needs replacement, Customer migration and transition management
DAY.AI FOUNDER
“(sales reps, recruiters, account managers, support staff) whose communications are recorded and analyzed, and”
Every company that sells anything needs a CRM. The current wave of AI has triggered a rush to "improve" existing CRMs by bolting AI features onto them.
The founders of today's startup – both veterans of HubSpot's product team – argue that approach won't produce anything genuinely transformative. So they built Day.ai: an AI-native CRM whose architecture was designed around AI from day one, rather than retrofitted.
Traditional CRMs made their users think like computers. First, design a schema: tables, fields, relationships. Then spend years trying to fit the messy reality of client relationships into that schema – manually or, more recently, with AI assistance.
But reality is always richer than our models of it. The result is an endless loop: either keep adapting the schema to match what's actually happening, or ignore the information that doesn't fit.
Day.ai skips the schema entirely. Users just communicate with clients through their normal channels – video calls, email, messages – and Day.ai figures out how to structure and store everything, building and adapting its own internal data models on the fly.
The platform is three products in one.
The first is People – an AI assistant that tracks all of a user's contacts. It auto-summarizes meetings, drafts follow-up emails, maintains a searchable contact history with context on every past interaction, and sends daily reminders about relationships that need attention.
The second is Pipelines – an AI assistant that surfaces potential deals from the communication stream. It auto-creates deal cards when a sales opportunity appears in a conversation, loops in relevant colleagues, keeps deal records updated after each new interaction, and maintains a live pipeline view with deal status and size.
The third is Pages – an AI assistant that builds the company's institutional knowledge base. It automatically surfaces recurring customer questions with video clips and verbatim quotes, compiles a ranked list of the most common product issues, creates per-customer issue logs, catalogs product feedback, and generates answer guides for frequently asked questions.
Users fall into two tiers: "Professionals" (sales reps, recruiters, account managers, support staff) whose communications are recorded and analyzed, and "Users" – everyone else who can view the knowledge base and deal cards and leave comments.
A Professional seat costs $50/month. Everyone else on the team uses the platform for free.
Day.ai is currently in testing, with integrations for Google Mail, Google Meet, Microsoft Teams, and Zoom – Slack coming soon.
Despite the early stage, Day.ai has raised $4M in a Sequoia-led seed round. The founders say they're not planning another raise anytime soon – the product funding is there, and after that comes revenue.
The driving logic behind Day.ai is the most interesting thing about it: instead of adding AI to an old architecture, replace the architecture with AI.
This is a broader pattern – rebuilding established product categories from the ground up using AI, rather than incrementally upgrading them. New products built this way can end up dramatically more useful than the originals, not incrementally better.
That shift creates a real opening. Incumbent products are built on old principles. Fully rewriting them is often harder than building fresh – and before you can even start, you have to admit the old version is fundamentally broken. Not every market leader is capable of that.
Motion, [covered here](/review/schaste-jeto-kogda-uspevaesh-delat-vazhnoe), did this for task management – an AI-native scheduler that automatically sequences work to maximize output and minimize missed deadlines, raising $63.6M after graduating Y Combinator.
[Yesterday's review](/review/uluchshat-nuzhno-otsjuda) covered Hark, which concluded that users would communicate with support teams more naturally by recording short video updates than by writing emails. But those videos would be unworkable without AI to extract the key information – the AI tools are what make the new input format viable. They raised $5M.
EdLight, [covered here](/review/uchit-nuzhno-na-salfetkah), built a platform that lets students learn by photographing handwritten math work on paper. The AI analyzes the handwriting. It raised $7.25M.
These are just the examples that come to mind quickly. The list is much longer.
The startups that win are those that catch a change happening right now – in markets, technologies, regulations, or user behavior. AI is exactly that kind of change.
But most startups respond by grafting AI onto products that are still structured the old way. The incumbents they're competing against can do that too – and have far more resources and existing customers to do it with. That's a hard race to win.
The different bet is rebuilding the product's foundations for an AI-native architecture – not adding to the old, but replacing it. When the result is qualitatively easier and more effective, the new player has a genuine shot at displacing the incumbent.
The upside: you don't have to discover a new market or validate demand. The old products already proved the need and defined the market. The only thing left to prove is that the new approach is meaningfully better – and offer a clear migration path.
What familiar product categories are ready to be rebuilt AI-native? What specifically needs to be replaced rather than added to? In what ways would the result be qualitatively simpler? Where does AI enable step-change improvements in effectiveness?
That's probably the most productive lens for targeting large, established markets and going after their incumbents right now.