LightTag

Hi! We Are LightTag

Our mission is to increase the velocity and yield of natural language processing projects. We execute on our mission by building products that isolate and remove constraints in the NLP workflow, for example annotation and quality assurance for labeled data.
We think that the infrastructure and best practices around labeling data are constraining progress in industry and academia. We also think that as a team, we are uniquely positioned to resolve those constraints. As NLP practitioners, these problems are important to us.

Tag Along, Join Our Team!

We are located in Berlin and in Tel Aviv and we are looking for excelent people to join our endeavour. Check these career options:
Senior Front End EngineerPixel Perfect Ninja!

We’re looking for a front end engineer who can deliver on both pixel perfect implementations of UI and high performance engineering of interactive web apps. We do interesting things here, you’ll work on recursive components for dependency tree visualization, parallel data processing with web workers and advanced full text search functionality with indexDB. The browser is the modern operating system and we want to leverage that to bring users the best possible experience in NLP.

The fact is, that algorithms aren’t enough, a great application needs to be beautiful, well thought out and flawlessly implemented. We’d like you to have a compulsion for beauty and order and see you insist on it both in the app and code you produce.

We aren’t concerned with a particular background, as long as you can do the work, can admit when you don’t know how and have the drive to figure it out. We’re going to support you, see you grow and have fun together. We’re a distributed team between Tel Aviv and Berlin, remote work is optional but encouraged though all of our offices are dog friendly. You can expect a competitive salary, a real equity stake in the business and plenty of greenfield opportunities to flesh out new ideas and put them in production.

Sounds interesting ? Here’s what the process looks like:
  1. Read this ECMAScript proposal and send a paragraph to hr@lighttag.io with a paragraph about the proposal and why you think we care about it.
  2. We’ll get back to you to set up a 45 minute call. During the call we’ll discuss your background and interests and talk a bit of code.
  3. If that works out, we’ll invite you to our offices for a 1.5 hour on site interview.
  4. And if that goes well, we’ll hire you for a week as a paid contractor so we can see what it’s like working together.
Head Of DevOpsMaster See It All

Our customers love us because LightTag just works and we want you to help keep it that way as we continue to scale. We run a multi tenant SaaS as well as on-premise installations, including air-gapped installations in highly regulated and secured environments. It will be your job to make sure things are working, and that if something goes wrong we know about it before our customers do.

You can expect to inherit a modern and well operating DevOps environment, one that ships code multiple times a day and where crises are few and far between. You’ll keep things that way as we triple our user base and deploy larger, more compute intensive AI models.

A signifiant part of the role is perfecting our processes around on-premise deployments. You’ll be meeting with enterprise customers, helping their tech team integrate our product and monitor it, while developing simple and repeatable processes for doing so at scale. In that sense, you’re not just comfortable with docker and aws, you’re also happy assisting those with less fortunate devops regimes get the most of LightTag.

We’re a distributed team between Tel Aviv and Berlin, remote work is optional but encouraged though all of our offices are dog friendly. You can expect a competitive salary, a real equity stake in the business and plenty of greenfield opportunities to flesh out new ideas and put them in production.

Sounds interesting ? Here’s what the process looks like:
  1. Send an email to hr@lighttag.io with a paragraph on why Kubernetes when the container orchestration wars.
  2. We’ll get back to you to set up a 45 minute call. During the call we’ll discuss your background and interests and talk a bit of code.
  3. If that works out, we’ll invite you to our offices for a 1.5 hour on site interview.
  4. And if that goes well, we’ll hire you for a week as a paid contractor so we can see what it’s like working together.
Lead NLP Engineer(Over) Fit Trainer

We make tools to make labeling data for nlp easier and faster. A major part of that is providing users with models that work seamlessly and actually help them work smarter and faster in tangible ways. Our customers work in many languages and domains, often with few existing resources and we typically can’t see their data anyway.

It’s your job to find ways to help them automate, whether through pre-training, secondary tasks or other things only you know about. It’s crucial that our users experience is top-notch throughout and so you are highly concerned not just with model accuracy but it’s throughput and training time.

Everyone loves deep learning and at LightTag we also love everything else that works. You should be a grand master of regular expressions, fluent in dependency grammars and of course, comfortable with an LSTM, a transform and a convnet (or some combination thereof). You love labeled data because it helps get things done, and getting things done matters to you more than the inherent beauty of the model you’re building.

That’s not to say we don’t value beauty and elegance, and we look forward to seeing what you come up with. Some of the recent NLP problems we’ve been working on are

  • What’s the strongest character level model we can in the users browser ?
  • Given a regular expression, how many distinct trigrams does it induce and can we enumerate them efficiently ?
  • Given the query history over some text dataset, can we partition it into disjoint subsets based on query results ?

Sounds interesting ? Here’s what the process looks like:
  1. Send an email to hr@lighttag.io with a paragraph on why are why not you should use BERT in production.
  2. We’ll get back to you to set up a 45 minute call. During the call we’ll discuss your background and interests and talk a bit of code.
  3. If that works out, we’ll invite you to our offices for a 1.5 hour on site interview.
  4. And if that goes well, we’ll hire you for a week as a paid contractor so we can see what it’s like working together.
Any other question or suggstion? please contact us!

LightTag make's labeling data
100% faster!
TEAM + AI
at scale easy
.