KICKSTART – Data Scientist – intern
Praksa Puno radno vreme Novi oglas

LLMs and agentic AI are changing how data science gets done, and we're looking for interns who want to be part of it.

The AI Strategy & Solutions team at Yettel is hiring three Data Science Interns, each with a slightly different mix of skills. Read the three tracks below and tell us which one fits you best. If more than one appeals to you, let us know that, too.

You'd be joining one of the region's largest telecommunications companies, where our data science teams turn data into products that customers use. From your first week, you'll be part of a cross-functional team, working alongside data engineers, AI specialists, and domain experts on code that makes it into production. We'll give you real data, real mentorship, and the room to grow.

Track 1: AI Engineering & Agentic Systems

You'll help design and build multi-agent systems that can reason, call tools, and hold a conversation. You’re familiar with agent frameworks, RAG, orchestration, observability and evaluation, and ideally speech pipelines (STT/TTS).

What you'll work on:

• Building and orchestrating multi-agent workflows
• Instrumenting agents for tracing, evaluation, and cost and latency monitoring
• Integrating speech-to-text and text-to-speech into conversational flows
• Prompt design, function/tool integration, and handling the failure modes of LLM-driven systems

Track 2: Machine Learning & Knowledge

You'll build numerical and machine learning models using large-scale, real-world data, turning customer and business signals into predictions that support decisions. You have strong statistical modeling skills, and ideally experience with MLFlow, knowledge graphs, and graph databases.

What you'll work on:

• Feature engineering and analysis on large-scale customer and business datasets
• Designing, training, and validating machine learning models
• Supporting models from development through production use, including experiment tracking, reproducibility, performance monitoring, and business impact measurement
• Turning entities and relationships in our data into knowledge graphs and graph databases (desirable)

Track 3: Model Training & the ML Platform

You'll run full training loops, end to end, on our multi-GPU AI platform. You have experience with machine learning frameworks such as TensorFlow, Keras, PyTorch, or sklearn, and ideally NLP and large-scale data curation.

What you'll work on:

• Generating synthetic data to fill gaps in coverage
• Curating and deduplicating a training-scale corpus across multiple GPUs
• Running a fine-tune and comparing it against the base model on consistent criteria
• Evaluating and optimizing model performance

For all tracks, you will bring:

• Final-year student or graduate in a quantitative, computer science, or engineering field.
• Proficient in both spoken and written English.
• Strong Python and SQL skills, with a good foundation in statistics and machine learning. Experience with PySpark and BI or data visualization tools is a plus.
• An interest in LLMs and where the field is heading, and the initiative to work things out independently.
• Strong organizational skills and knowledge of modern software development methodologies.
• You don't need every one of these. If you're strong in some areas and keen to learn the rest, we'd still like to hear from you.

Why we think you should apply:

• Paid 9-month internship with career development opportunities
• Mobile phone and tariff package with unlimited internet.
• Opportunity to participate in the business decision-making process of a large multinational company.
• An environment where you can freely express opinions and ideas, experiment with new AI tools, and inspire new ways of doing things.
• A unique and inspiring atmosphere full of talented people willing to share their professional knowledge.

Please include your references in your CV, as reference check is an integral part of our selection process.

We hope to meet you soon!

Označeno kao: 3 izvršioca