Tonar is recruiting an ML Engineer. This is a startup, before reading more about the job take note and be excited or discouraged by these statements:
- You will be part of a small team
- Your work can make it or break it for the company
- You are expected to ship features before they reach 100 % accuracy
- This is no place for ego, best solution wins, we move fast
- The salary for this role is 75.000-95.000 SEK/month, depending on seniority
- The role comes with stock options between 0.25-1.0 % of the current valuation
- Because of our current market, this role requires you to be
fluent in Swedish
.
-- About the company --
Tonar is changing the possibilities of media analysis. Our business is to find and index all media within the spoken word (radio, tv, social media, podcasts etc) and make it consumable for our users, both as real-time monitors and for deeper analysis. Typical questions we want our data to answer is:
- What's the latest news about Donald Trump?
- Analyze the sentiment trend of McDonalds over the past 6 months.
- How did Apple perform compared to Microsoft with regards to AI in the past year?
- Which Swedish politicians are supportive of stricter immigration laws?
ML is at the core of the company and you can expect to work with both transformers and smaller models in the areas of:
- Named Entity Recognition and reconciliation
- Graphs (who/what is/was related to who/what at time X)
- Speech to text (in multiple languages)
- Video understanding
- Massive scale content classification
- Relevance/ranking
- Agentic action systems based on user interaction
-- Responsibilities --
As a ML Engineer in the team you are expected to drive model development for features in the product together with the current AI/engineering team in close collaboration with the CTO and the CPO. Expect to be working on things where there is no solution online.
-- Qualifications --
Minimum
(seriously don't apply if this does not match you):
- Python skills in data science - Numpy, Pandas, Polars, Scikit, Huggingface, Pytorch, Tensorflow (whatever your stack is).
- You have at least a couple of years industry experience of at least 1 of the core areas we are focusing on (stated above).
- You have put models into production and have an understanding of the difference between a proof-of-concept and actually running it live.
- You have proper experience in AWS or GCP.
Desired
extras (if you have please state it clearly in your resume):
- You have experience with language models for smaller languages (nordics etc), specifically improving accuracy.
- You have experience fine-tuning writing production-grade LLMs prompting/handling.
- You have experience optimizing models from base to quantized etc. to improve the cost/performance.