Technical Innovation
Is at the Core of What We Do

Grammarly exists because our engineers have the freedom to innovate and uncover breakthroughs. These breakthroughs regularly influence the product roadmap and help chart the course for the future of our company.

Today, our full-stack team is aggressively scaling our interfaces and distributed computing (NLP, ML) to support 20M daily active users. The complexity of our technical challenges is growing exponentially, and we’re excited about growing and diversifying our team.

Recent Projects
One of our goals is making Grammarly widely available on different platforms, including web browsers, mobile devices, MS Office, and more.
Applied Research
We invest in a combination of NLP and advanced ML technologies to deliver actionable writing feedback to Grammarly users.

Browser Extensions

Grammarly's browser extensions offer direct feedback on writing in text fields. We use cutting-edge front-end stack. TypeScript ensures that we write type-safe, bug-free code. Focal, our open-source state management library, on top of Rxjs allows us to use a functional reactive approach. Webpack glues everything together. Learn more in our article "Building Browser Extensions at Scale."
Sergey Yavnyi, Tech Lead
I think what we’re doing in the browser extension space is pretty unique. There’s a lot of space for innovation, and the technical challenges are demanding. Because of this, we need to be very open and at the same time very disciplined when we’re making technical decisions. There is no known best practice to accomplish whatever we need for the product — we have to find it on our own.”
Join the Extensions Team
We have an open position for someone who has experience with low-level browser APIs and a love for type-safe languages.
Check Out Focal
We've built a type-safe, expressive, and composable state management library for React applications.

Mobile Keyboard

Writing well on mobile is often challenging and inconvenient. We built the Grammarly Keyboard to bring effective and mistake-free writing to mobile devices. Since launching our iOS and Android apps in 2017, we’ve gained incredible traction — our users are writing with the keyboard more than 60 times per day.
Dima Smolin, Senior Developer
The most challenging part here is dealing with API limitations and resource constraints for custom keyboards. Fitting into this restricted environment and still providing a good user experience requires many clever tricks and complex workarounds. Another challenge is that despite having a lot of functionality to expose, the keyboard cannot take up too much space on the screen. Everything, including the NLP code running on the device, the communication with servers, and all the other moving parts, needs to work really fast to respond swiftly as a user types. It gives you lots of opportunities to express your engineering skills and creativity.”
Join the Mobile Team
We are looking for experienced Android engineers to help us deliver world-class user experiences.
Tech Stack

Android: Java, Mockito, Retrofit, Guava

Writing Feedback

More than two billion people worldwide use English in their personal and professional communication. Whatever their purpose, most people care that their writing is polished and conveys their ideas well. This is where Grammarly checks come in. Our team of computational linguists and research engineers develops complex algorithms for detecting and correcting errors, ambiguities, and inefficiencies in natural language texts. Grammarly checks billions of words every day, ensuring our users’ writing is clear, effective, and error-free.
Igor Tytyk, Research Engineer
There are multiple ways to write a sentence — and just as many ways to make a mistake in it. Given the fascinating and irregular nature of human language, the task of natural language processing is as exciting as it is challenging. At Grammarly, we combine the latest machine learning approaches with a healthy dose of human ingenuity to provide valuable and accurate writing assistance. Plus, it’s motivating to work on the cutting edge of recent scientific innovation to help native speakers and language learners write better.”


To process natural language texts properly, we need to understand how the language functions, how it is learned, and how it evolves. Syntactic and semantic parsers as computational linguistics tools allow us to extract structured information from millions of pieces of raw text. Parsing is a key part of our text processing pipeline, and the linguistic structure it produces allows Grammarly to provide real-time writing feedback.
Mariana Romanyshyn, Computational Linguist
Every piece of text that passes through Grammarly undergoes a number of text processing steps, from splitting into paragraphs and language identification to syntactic parsing. Like any other natural language processing tool, syntactic parsers are subject to language ambiguities, so we constantly improve the data and techniques used. As a result, our language parsers are helpers for our error-correction and text improvement algorithms.”
Join the NLP Team
We are looking for people who have a passion for applied NLP research. Whether you like developing rules, designing ML classifiers, or building complex neural networks, we have a position for you.
Dependency Parse Tree Visualization

Impossible Is Nothing for Our Team

When Grammarly was founded in 2009, the concept of using AI to facilitate communication was a completely new concept. With a readiness to adopt new technologies, and the capabilities to customize open-source tools for our own use, we managed to build something profitable from the start.

We were originally told that developing a cross-browser extension that would work across the web would be impossible. Our browser plugin now works in all major browsers (Chrome, Safari, Firefox, and even Edge), and this was done with just a handful of engineers.

A lot of what we do hasn't been done before. Engineering innovation and creativity help drive us. That's why we've invested in full-day hackathons for engineers to prototype and develop new ideas, and it's why we set aside time to explore new tools and technologies, read up on industry trends, or hack up new widgets just for fun.