Division: Group Digital Capabilities (GDC)
The Group Digital Capabilities (GDC) Division ensures Euroclear’s competitiveness by delivering reliable and sustainable IT solutions for the financial securities markets.
Our technical teams deliver new IT solutions and improve existing applications for both our internal and external clients. We deploy changes into the production environment in a controlled and structured way that does not compromise production stability, and we ensure applicative production support.
Our non-technical people maintain the maturity of the IT project delivery with appropriate controls in line with the group’s risk appetite and reducing development and running costs.
Within the GDC Division, the AI CoE supports the Data Science needs of all the Euroclear Group entities. As a competency centre for AI/ML, the team helps improve process efficiency and generate insights using techniques such as predictive modelling, natural language processing, GenAI and mathematical optimization.
Qualifications
Demonstrable record of hands-on experience in the area of AI/ML/Advanced Analytics, with special focus on deploying and maintaining AI/ML models and services in production.
Keywords: AI/ML application development, testing, serving, monitoring, troubleshooting.
Knowledge on how to ensure ML models are reproducible and interpretable.
You have already single-handedly packaged and deployed AI/ML services to production.
Ability monitor and maintain AI/ML services post-deployment.
Proficient in Python
Has more than 5 years of experience with Python, and AI/ML standard libraries such as pandas, scikit-learn, xgboost
Nice-to-haves:
Data processing libraries and frameworks (pydantic, pandera)
Web frameworks (such as FastAPI, Flask)
CLI frameworks (Typer, Click)
General MLOps tools and frameworks (MLFlow, Azure ML Studio)
Version control tools for ML datasets and models (DVC, Azure ML Dataset)
Monitoring libraries and solutions (such as NannyML, Evidently AI)
Distributed processing libraries and frameworks (such as Ray, Dask, PySpark)
Pipeline-building and orchestration libraries (such as Metaflow, ZenML, Kedro, Airflow, Dagster)
General Python development tool (pytest, coverage, tox, mypy, black, ruff, uv, pip-compile)
Ability to write both object-oriented and functional code, and understand concepts such as (de)coupling, coherence, inheritance, composition.
You make sure the code that you and your colleagues write is thoroughly tested (unit, integration, end-to-end, stress/performance).
You love and regularly use data validation and type hints.
You know how to turn a messy jupyter notebook into a production-grade piece of code.
Although we'll apply all possible preventive measure to prevent this from ever happening.
Knowledge on how top package a python application or library for distribution
Proficient GIT user, able to collaborate with multiple developers on multiple repositories, while following standard processes related to branching, merging and code reviews.
Good understanding of Machine Learning algorithms and their applications in NLP.
Work experience with at least one Cloud Provider, preferably Azure Cloud.
Experience with Unix/Linux command line tools and scripting (shell, bash):
VIP club membership if you have at least once ran `rm -rf` on production data.
You possess the foundational Data Engineering skills, allowing you to interact with the Data Engineering team, and analyze and troubleshoot data pipelines if needed:
You could handle using SQL to extract, transform and load data (ETL/ELT).
Experience with the Hadoop ecosystem (Spark, Kafka, Hive, Impala…) is a plus.
Experience with the Cloudera distribution is an additional plus
Understanding of modern MLOps framework and complexities it adds to DevOps.
Ability to identify the MLOps maturity gaps and provide inputs for modernization efforts.
Non-technical
Strong verbal and written communication skills as well as good customer relationship skills to present sophisticated concepts and/or the results of a use case to different audiences (from end users up to division management).
Experience of working in large, sophisticated enterprises and have stoically accepted it as your fate.
You are not allergic to legacy technology, yet are always on the lookout for modernization opportunities.
Stay up-to-date with new tools, technologies and approaches within the domain.
Well-integrated standout colleague.
Able to estimate your short-term effort with reasonable accuracy and get the work done in the time frame you commit to.
Optimally swim in the waters of Agile project management techniques (scrum boards, standups, demos, reviews).
Stand to promote MLOps and advocate for its usage and necessity across the organization.
Must love mentoring and sharing knowledge!
Must love dad jokes!
Your formal qualifications are the following:
University degree in software engineering OR Data Science/Machine Learning/Data Engineering OR a related quantitative field, combined with strong IT skills.
More than 5 years of experience with Python
More than 2 years of experience of using DevOps/CI/CD practices and deploying AI solutions to production.
#LI-ME1
