Senior Machine Learning Engineer at at Rainbow Weather (Part-time work/Remote)
We are a start-up called "Weather Forecast Technologies - Rainbow.app" and we are building the most accurate weather service in the world Rainbow Weather. Our team collects data from a wide range of sources (radar, satellites, barometers, etc) and applies ML models to it, making an innovative product. Our solution includes: web, android, ios client applications, a technologically diverse back-end, several different AI/ML services, monitoring and metrics collection. You must have seen us as winners of the day and week at Product Hunt last August and the other day as nominees for the Golden Kitty Awards in the best mobile app category.
If you don't know about us, you can listen to thepodcastwith investor Yura Melnicek.
Over the last year of operation, we have done
By metrics we make the best precipitation forecast two hours ahead with updates every 10 minutes.
By metrics in North America and Britain we get ahead of Apple and worldwide we got ahead of AccuWeather and of The Weather Channel
We made an app that became successful on Product Hunt in the Golden Kitty Awards award in the best mobile app category.
What makes us different from other apps? We have cool features that make us better than others. For example, push notification for sudden weather changes. If it rains in your location, we'll make sure to let you know to bring an umbrella.
Our goals for 2024
The best minute-by-minute weather forecast in the world for 24 hours with updates every 5-10 minutes.
We want our app to operate covering the entire globe
Strengthening the development of our application and separately its B2B direction
We are now looking an excellent Senior Mahine Learning Engineer!
Independent organization of their work in the tasks of the product
Participation in large meetings once a week and daily stand-ups
The manifestation of initiative and its argumentation
Finding the best solutions for continuous product improvement
Development of state-of-the-art deep learning based weather prediction models:
From early prototype to production
On the global scale
In automated manner, utilizing AutoML approaches, so each region have their best model
From nowcasting (3 minutes) to hours and days
Successful collaboration with other team members
Deep knowledge of convolutional neural networks: architectures, losses, typical tasks (segmentation, classification, object detection)
Basic understanding of recurrent neural networks (LSTM, GRU, Transformers) and time series analysis
Basic understanding of GANs
Hands-on experience with modern frameworks - PyTorch, TensorFlow
Experience in delivery of production-ready ML solutions
Basic knowledge of statistics
Ability and openness to learn, fail, reflect and apply learning moving forward
High level of self-organization
Ability to solve complex engineering problems in favor of business problems
High degree of involvement in the product
The ability to develop the product in depth
Would be nice to have
Experience working with satellite imagery
Experience working with RADAR/LiDAR data (or any other sparse types of data)
Experience with AutoML approaches
Experience with anomaly detection techniques
What we offer
We offer solid stock options compensation for the result. It's a chance to earn much more then regular salary.
We can help with relocation to EU if you need it.
We have comfortable office in center of Warsaw to offline meeting, the company will cover all transportation costs for this meeting.