Data Scientist
Data scientist at telecom operator Yettel working at the intersection of applied AI and telecommunications: designing multi-agent LLM systems, training and fine-tuning models, and taking them to production. Core stack is Python, SQL, classical ML, and LLM/agentic tooling.
- Compensation
- Not disclosed
- Location
- Beograd, RS, 11070
- Work setup
- Not specified
- Seniority
- Not specified
- Employment
- Not specified
- Experience
- Not specified
- Category
- Data science
- Posted
Skills
Must have
- Strong experience building ML or AI systems, with a track record of taking work into production
- Good grounding in classical machine learning and hands-on experience with LLMs
- Solid mathematical and statistical foundation
- Fluency in Python and SQL and the ML/data science ecosystem
- Clear communication and judgement when problems are ambiguous
- Experience mentoring others or leading projects end to end
- Background in telecommunications or similarly large-scale, real-world data
- Business acumen and ability to communicate technical developments in business terms
- Include references in CV; reference check is an integral part of the selection process
Full description
Are you excited by how LLMs and agentic AI are reshaping the way we build with data? We're looking for a data scientist who wants to work at the intersection of applied AI and telecommunications: designing agentic systems, training and fine-tuning models, and taking both from idea to production on a real platform with real users. You'll join a cross-functional team of data scientists, engineers, and domain experts, and you'll have the data, compute, and autonomy to do meaningful work and to help shape how we do it.
What we will expect from you:1) Build agentic AI systems
- Design and build multi-agent systems
- Own orchestration, observability, and evaluation so these systems are reliable, measurable, and improvable
- Work through the hard parts of LLM-driven software: prompt and tool design, latency and cost, and anticipating failure modes
2) Train and improve models
- Run full training and fine-tuning loops
- Curate and deduplicate training-scale datasets, and generate synthetic data to fill gaps
- Evaluate models and close the loop by feeding production signals back into the next training cycle
3) Take work to production
- Move models and systems from prototypes into reliable, maintained services
- Collaborate with data and ML engineers on the pipelines, evaluation, and monitoring that keep them running
- Help raise the bar on quality and share what you learn with the people around you
For a good start you will need:
- Strong experience building ML or AI systems, with a track record of taking work into production
- A good grounding in classical machine learning and hands-on experience with LLMs
- A solid mathematical and statistical foundation, with fluency in Python and SQL, and the ML / data science ecosystem around them
- Clear communication, and the judgement to make good calls when problems are ambiguous
- Experience mentoring others or leading projects end to end
- Background in telecommunications or similarly large-scale, real-world data
- Business acumen and ability to communicate technical developments in business terms
Why we think you should apply:
- Working in a dynamic and friendly environment which allows you to grow and develop
- Flexible working hours with the possibility of working remotely
- 25 days off per year
- Competitive salary and participation in the bonus reward system