The quiet API plumbing under every modern inbox, calendar, and meeting.
Somewhere right now, a recruiter is reading a candidate reply, a salesperson is auto-logging an email into a CRM, and a patient is booking a telehealth visit through a calendar that magically syncs back to a doctor's outlook. None of them are thinking about Nylas. That is the whole pitch.
Nylas is a San Francisco software company that sells one product to people who do not enjoy talking about email: a single API that connects any application to any email, calendar, and contacts provider. Gmail, Microsoft 365, Exchange, iCloud, IMAP, the regional provider in a country you cannot pronounce - all of it, behind one HTTP endpoint and a normalized data model.
The numbers are bigger than the brand recognition. More than 250,000 developers have built on the platform. Hundreds of companies - Upwork, Wix, Freshworks, Lever, Dialpad, Move.com - use Nylas to do something their users will never see and never thank them for: stay in sync with the inbox. Total funding sits at $175 million, the last $120 million of it written by Tiger Global in 2021.
Email is fifty-something years old. Calendar protocols are not much younger. Both were designed before anyone imagined a SaaS app would need to read, send, write, and reason about them on behalf of millions of users. The result is a stack of specifications - IMAP, SMTP, MIME, CalDAV, CardDAV, EWS, Microsoft Graph, the Google APIs, OAuth in five mildly incompatible flavors - that any ambitious application eventually has to wrestle with.
Wrestling, here, is generous. It is closer to negotiating with a parliament of strangers, each of whom has opinions about timezones.
For a long time the industry's answer was, essentially, hire people. Big companies built integrations teams. Smaller companies built one provider, hit "ship," and quietly hoped no enterprise customer would ask for the next one. Nylas's bet was that this work was so universally annoying, so universally avoided, that a layer of infrastructure could do it once - and rent it back to everyone.
Maintain OAuth tokens for six providers. Handle nine flavors of MIME parsing. Translate "free/busy" across calendars that disagree on what "free" means. Retry, dedupe, throttle. Re-do all of it when Microsoft changes its API. Pay someone to read the IMAP RFC at midnight.
Nylas was founded in 2013 by Christine Spang and Michael Grinich, two MIT alums who started by trying to build something far more romantic - an open-source email client called Inbox - and ran face-first into the same fragmentation everyone else hits. Spang had previously spent her early career hacking the Linux kernel at Ksplice (later acquired by Oracle), which is to say: she is comfortable in the parts of the stack that other people quietly route around.
The pivot from "build a client" to "build the infrastructure other clients need" is one of those moves that looks obvious in retrospect and felt, at the time, like admitting defeat. It was a good piece of admitting defeat. Within a few years, the team had a clear thesis: communications data is the new database. If a company could become the canonical access layer for it, the rest of the API economy would route through them.
Nylas sells four things that all amount to the same thing: a way to skip the boring half of your roadmap.
Send, receive, search, thread, label, and sync mail across every major provider with one consistent schema. Webhooks tell your app when something changes, so you stop polling like it is 2009.
Read and write events across Google, Microsoft, iCloud, and friends. Free/busy queries work the way you wish they did. Time zones are handled by adults.
A unified contact model that does not blow up the second a record contains an emoji or a non-Latin name. Reads, writes, dedupes.
A drop-in scheduling experience (think Calendly, but inside your product, with your brand) and, more recently, an AI Notetaker layer that records, transcribes, and summarizes meetings using the same provider graph.
An infrastructure company is judged by who quietly uses it. Nylas's customer list reads like a directory of categories you'd expect: recruiting (Lever), freelance marketplaces (Upwork), website builders that grew into businesses (Wix), customer experience platforms (Freshworks), unified communications (Dialpad), real-estate listings (Move.com). The pattern: products where email and calendar are not the feature, but the feature does not exist without them.
Officially, Nylas wants to turn communications infrastructure into a programmable resource any developer can plug into in a few minutes. Unofficially, the mission is something closer to: spare the next generation of builders the indignity of debugging an OAuth flow at 2am.
The internal values - Trustworthy, Playful, Nurturing, Unstoppable - are listed on a careers page in that order, which is unusual. Most infrastructure companies put "Unstoppable" first. Putting "Trustworthy" at the top, for a company whose entire job is to handle other people's email, is the only sensible thing.
It is fashionable to believe that AI will rewrite communications software from scratch. Maybe. But large language models still need somewhere to read, write, and act. They need calendars to book against and contacts to address and inboxes to send from. The boring graph underneath the agentic future looks, suspiciously, like Nylas.
The next ten years of software will not be a fight over inboxes. The inbox is settled. The fight is over what sits on top of communication data - the AI assistants, the workflow tools, the vertical SaaS for healthcare and real estate and recruiting that need to act on messages and meetings as fluently as they act on rows in a database.
For any of that to work, someone has to keep the pipes clean. Someone has to handle the OAuth renewals, the throttling, the encryption-at-rest, the new Microsoft Graph endpoint shipped on a Tuesday. That role has, almost by accident, become Nylas's.
Back to the recruiter, the salesperson, the patient. None of them know the name of the company connecting their software to their inbox. None of them will. That's the deal Nylas made a decade ago, on the day Spang and Grinich decided the email client was a bad business and the infrastructure underneath it was a great one. They were right. The proof is the fact that you stopped noticing.