Senior Data Scientist

Currently, we are looking for a Data Scientist who will be a part of the general-purpose data science team. We define a Data Scientist (DS) as a scientist(!) as opposed to a data analyst with fancy modeling that is able to formulate hypotheses, arrange experiments and adjust modeling approaches to the problem at hand. Though we expect that a senior data scientist has close ties with product vision and is able to overview problems from the business perspective.

In this position, you will often communicate with a customer and consult both technical and non-technical team members regarding questions within your domain of machine learning and data-driven solutions.

Responsibilities:

Understand business needs and restrictions and offer appropriate technical solutions in the domain of data-driven applications, describe qualities and limitations of proposed approaches to non-technical people.
Work with all types of data including tabular, visual and textual data, state data collection and labeling requirements.
Choose, design and implement data modeling techniques from heuristics to classical machine learning to deep learning models.
Research SOTA approaches, set up experiments for hypotheses validation and improve existing modeling techniques.

Requirements:

Fundamental (must have 80% of these)

Knowledge of linear algebra and calculus sufficient for explaining concepts such as backpropagation and regularization.
Knowledge of fundamental machine learning theory such as soft optimization and statistical learning.
Knowledge of classical machine learning algorithms such as ensembles, clustering and dimensionality reduction.
Knowledge of basic deep learning architectures such as MLPs, CNNs, RNNs, and, maybe, a little bit of Attention.
A brief touch to theoretical computer science and its algorithms (we don’t ask inversion of binary trees).
Experience with basic software engineering instruments such as python, bash, git and some relational databases.
Understanding of Data Science Lifecycle from data collection to model training, deployment and maintenance.

Domain (having at least 3 points in total is already great)

Building production-ready computer vision systems including camera and infrastructure requirements.
Handling deep learning solutions from Detection to Tracking to Segmentation or Pose Estimation.
Dealing with classical approaches for Tracking, Trajectory Estimation and 3D Scene Construction.

NLP

Knowledge of theoretical concepts in Natural Language Processing such as encodings, embedding spaces and language heuristics.
Commercial usage of transformer-based models for common tasks from Translation to Named Entity Recognition to Language Modeling.
Making large-scale language models work in production such as BERT, T5, GPT and any other flavours of these.

Building recommender systems based on collaborative filtering and content matching, ideally based on deep learning.
Dealing with time-series such as market values or resource allocation with predictive models and anomaly detection.
Solving analytical problems with explainable pipelines based on clustering, explainers and white-box modeling.

Scaling up (a few ways to stand out)

Ability to read, estimate and implement solutions from scientific papers without code provided.
A number of somewhat cited publications in scientific journals or conferences.
Lecturing and mentoring experience in an either academic or industrial setting.
A fancy scientific or engineering degree related to computer or data science. (MS+)
Fluent knowledge of English or any additional language across four major skills. (ideally C1+)

You will work with smart people who love to solve hard problems, and who not only expect but also foster high performance!

If you fit the description above, we’d love to hear from you! Email us at hrm@indatalabs.com.