Curriculum
Admission Process
Community
FAQ
Become an expert
Starts March 17
Applications are closed

AI Performance Engineering

Gain practical AI engineering skills, learn LLM internals and deployment in 14-week hybrid program at Tel Aviv University campus.

Computer screen displaying Python code with tabs labeled Preview, Code, and Blame, alongside floating icons of a code bracket and a Git branch symbol.Code snippet showing a Python function named eval_agent with tabs for Preview, Code, and Blame.Code snippet showing a Python function eval_agent that builds a container, runs an agent, collects a patch, and evaluates the patch by applying it to a test container.
AI basics to production

From LLM internals and training to scalable deployment, MLOps, and performance .optimization

Learn real AI workflows

Inference scaling, orchestration, RAG systems, experiment management, and .post-training

Build real-world projects

Work on real projects that focus on practical AI methods and .hands-on experimentation

Learn from top experts

Study with engineers and researchers from Nebius, AI21 Labs, Deepchecks

A structured, intensive
14-week program

Hybrid learning at Tel Aviv University

Lectures take place on campus at Tel Aviv University and are streamed live online for full remote participation

Group of graduates smiling and holding Nebius Academy diplomas during a graduation ceremony.Group of graduates smiling and holding Nebius certificates at Nebius Academy Graduation Day.
Duration: 14 week
Workload: 85 hours of classroom study and over 280 hours of homework
Pace: 2 classes per week + bi-weekly home tasks
Group of graduates smiling and holding Nebius certificates at Nebius Academy Graduation Day.

Who is this for?

Solution architects working with cloud-based AI systems

Software developers & engineers building or integrating AI features

ML engineers, data scientists & DevOps specialists involved in model deployment and tooling

Professionals needing to quickly launch AI workflows and grasp LLMs in production

Prerequisites

Confident Python programming experience
Basic ML knowledge is welcome
Familiarity with cloud platforms or deployment tools is a plus
Join the Waiting List
Audience seated in blue chairs clapping and smiling during an event or presentation.Audience seated in blue chairs clapping and smiling during an event or presentation.

Hands-on learning

Weekly practical labs

From LLM internals and training to scalable deployment, MLOps, and performance optimization

Home tasks

Bi-weekly structured home tasks

Practical capstone

Capstone projects focused on building and deploying production-grade AI systems.

Real experimentation

Access to tools and compute for real experimentation

Program structure

Module 1, March 17–May 8, Tuesdays 17:30–20:30
From AI model to AI product

How to build AI agents for your business tasks on top of a third-party API
Module leader: Yuval Belfer (AI21 Labs)
Guest Lecturer: Tom Braude (Stealth Startup)

Module 2, March 20–May 26, Fridays 09:30–12:30
LLM Architecture

From neural network essentials to the architectural choices that shape today's LLM landscape
Module leader: Noa Yehezkel Lubin (Israel AI Agency)

Module 3, May 12–June 16, Tuesdays 17:30–20:30
MLOps

Production deployment and scaling of AI systems, MLOps tools, experiment management, and monitoring
Module leaders: Simon Karasik (Nebius) & Liran Jdanov (Nebius)

Module 4, June 5–July 10, Fridays 09:30–12:30
AI Performance Engineering

LLM performance engineering: from speculative decoding and quantization to custom CUDA kernels
Module leader: Asaf Joseph Gardin (AI21 Labs)
Guest Lecturers: Yonatan Glassner (NVIDIA), Yehoshua (Shuki) Cohen (AI21 Labs)

Module 5, June 23–July 7, Tuesdays 17:30–20:30
Post-training & RL

Post-training and alignment of LLMs: from SFT to RL and back
Module leader: Elik Sror (WSC Sport)

Certification

Students receive an official Nebius Academy AI Engineer Certificate upon completion

Submit your application for the course right now

Apply now and start your path to becoming an AI Performance Engineer

Screenshot of a data science project interface showing files objective.md, FastDrop.csv, delivery-delays.ipynb, and a preview of FastDrop.csv with delivery person IDs and codes.Screenshot of a code editor showing files for a data scientist project: objective.md, FastDrop.csv, and delivery-delays.ipynb, with a snippet of FastDrop.csv displaying a table of IDs and delivery person codes.

Admission process

Through a careful selection process, we ensure a motivating and supportive learning environment and make sure every candidate has the time and ability to succeed in this intensive course

1
2
3
Application

Fill out the form to apply. You’ll get an email with details about the test and next steps.

Online Test

Take an online test assessing your analytical and programming skills

Interview

Meet the team in person or online to discuss your background, motivation, and goals; some technical questions may follow.

Acceptance to the course

Successful candidates will receive an offer and confirm their spot, starting their AI engineering journey.

Syllabus

5 Modules
Module 1: From an AI model to an AI product
5 weeks, 3h/week
Module 2: LLM architecture
8 weeks, 3h/week
Module 3: MLOps
6 weeks, 3h/week
Module 4: Performance engineering
6 weeks, 3h/week
Module 5: LLM post-training
3 weeks, 3h/week

Showcase your AI engineering
skills with credibility

Prove your expertise

Receive an industry-recognized certificate that validates your skills and completion of the program

Boost your professional credibility

Stand out to employers with a credential designed together with cloud & AI experts

Showcase your achievement

In one click, add your certificate to your LinkedIn profile, CV, and portfolio to highlight your professional growth.

We help you turn AI capabilities into innovation

As Nebius's education and research arm, Nebius Academy helps professionals and teams apply AI in real workflows and support innovation.

Smiling man and woman standing closely together in a cozy, modern café setting with shelves and glasses in the background.Young man with short curly hair wearing a striped shirt and blue lanyard looking intently to the right.
15,000+
learners across Europe
the US, and Israel
3,000+
engineers, researchers, and tech leaders in our community
Audience seated in a theater clapping and smiling during an event.Woman with brown hair in a bun, wearing glasses, a maroon shirt, and a blue lanyard, standing indoors near a wooden wall and a blurry poster.
Smiling man and woman standing closely together in a cozy, modern café setting with shelves and glasses in the background.Young man with short curly hair wearing a striped shirt and blue lanyard looking intently to the right.
15,000+
learners across Europe
the US, and Israel
3,000+
engineers, researchers, and tech leaders in our community
Audience seated in a theater clapping and smiling during an event.Woman with brown hair in a bun, wearing glasses, a maroon shirt, and a blue lanyard, standing indoors near a wooden wall and a blurry poster.
Smiling man and woman standing closely together in a cozy, modern café setting with shelves and glasses in the background.Young man with short curly hair wearing a striped shirt and blue lanyard looking intently to the right.
15,000+
learners across Europe
the US, and Israel
3,000+
engineers, researchers, and tech leaders in our community
Audience seated in a theater clapping and smiling during an event.Woman with brown hair in a bun, wearing glasses, a maroon shirt, and a blue lanyard, standing indoors near a wooden wall and a blurry poster.
Smiling man and woman standing closely together in a cozy, modern café setting with shelves and glasses in the background.Young man with short curly hair wearing a striped shirt and blue lanyard looking intently to the right.
15,000+
learners across Europe
the US, and Israel
3,000+
engineers, researchers, and tech leaders in our community
Audience seated in a theater clapping and smiling during an event.Woman with brown hair in a bun, wearing glasses, a maroon shirt, and a blue lanyard, standing indoors near a wooden wall and a blurry poster.

Alumni testimonials

Israel has been a strategic region for Nebius since 2018. Prepared hundred of top specialists working in the top companies in the Industry

Rachel Shalom

Job: Principal Data Scientist, Dell Technologies

Course: AI-Powered Data Science

(formerly Y-DATA)

I realized that as a product manager in a travel tech startup, I needed heavy tools to analyze data, do predictions and more. So I started checking all kinds of data science boot camps, and machine learning academies, but unlike most of them, Y-DATA looked realistic. I chose Y-DATA because one year is better in terms of understanding things. Also, I could combine it with my previous work.

Yael Hamrani

Job: Data Science & Integration R&D Engineer

Course: LLM Engineering Essentials

The course is practical, insightful, and the community aspect adds a lot of value. I’d recommend it to anyone looking to meaningfully expand their AI skill set.

Dina Karakash

Job: Venture Partner and Head of AI & Research

Course: Intro to ML from an LLM Standpoint

Most students were already in the field, tackling real AI challenges — from building models and AI agents to applying flexible AI frameworks in specific industries. The lecturers were both researchers and practitioners deeply passionate about AI. This course went well beyond 36 hours of content.

Ido Nissim

Job: Data Engineer at AllCloud

Course: AI-Powered Data Science

(formerly Y-DATA)

I think the very best thing about the course is the people. The selection of the students for the course was really good. Heterogeneous people from all kinds of fields and different backgrounds — that’s really good. We had some projects together, and worked as groups, which was a good way to get to know other people. We were all sitting in the classroom together, talking and trying to figure out how to do the homework later on. It’s great.

Andi Mardinsyah

Job: Data Scientist

Course: Practical Generative Al

I chose the Generative AI program from Nebius Academy because of its well-structured curriculum that balances theory and practice. The first module on Generative AI applications was extremely useful, especially with its hands-on coding approach. Despite my busy schedule, the short yet informative lectures made learning manageable. This program is a great fit for Data Scientists working with NLP and LLM, and I’d definitely recommend it to my colleagues.

Liad Yosef

Job: Principal Software Engineer, Shopify

Course: AI-Powered Data Science

(formerly Y-DATA)

You know they say go with your passion, right? I’ve been programming since I was a kid, but I never really dealt with Data Science or Machine Learning before Y-Data. I already knew the math part of the introductory courses but they were so fast-paced that I wasn’t bored and quickly enough we got into supervised learning and deep learning. This gave me the tools to do things that I couldn’t have done before, and let me explore and widen the area of my thoughts.

Tal Heletz

Job: Deep Learning Researches at Trigo

Course: AI-Powered Data Science

(formerly Y-DATA)

It was obvious to me that math is the field for me. I did my B.Sc and M.Sc in math. In the industry, you can do a lot with math, but you must have knowledge in computer science as well. Y-Data was exactly right for me — it let me combine my background with computer science and strong data science foundations.

Andrey Nikitin

Job: Data Science Manager, Cyera

Course: AI-Powered Data Science

(formerly Y-DATA)

The course is great, I think it’s the best professional course I have taken and for me personally, it’s a good substitution for a master’s degree (for now). Even though I’m already working as a Data Scientist i still learn new things, there are always fields that I’m less proficient in and the course fills the gap.

Faculty & mentors

Smiling young woman with long dark hair wearing a beige shirt against a light gray background.
Noa Lubin
VP Technology
at Israel AI Agency
Read more
Noa Lubin is VP of Technology at the Israel AI Agency and an AI leader with extensive experience in NLP, healthcare, and space technologies. Previously Director of Data Science at Fido, she has held roles at NASA, Amazon, Diagnostic Robotics, Elbit Systems, and the Israeli Aerospace Industry, leading and applying machine learning solutions in high-impact domains.

Noa holds an MSc in Computer Science (Magna Cum Laude) from Bar-Ilan University, where she specialized in Natural Language Processing under the supervision of Prof. Yoav Goldberg, and a BSc in Electrical Engineering (Summa Cum Laude) from the Technion. Alongside her industry work, she is an active lecturer and public speaker in AI and data science, combining deep technical expertise with a strong commitment to socially impactful AI initiatives.
Smiling man with short dark hair and beard wearing a black deepchecks t-shirt on a dark background.
Philip Tannor
CEO
at Deepchecks
Read more
Philip Tannor is the Co-Founder and CEO of Deepchecks, an ML and LLM evaluation company known for its open-source validation tooling and practical approaches to monitoring and quality measurement in production AI systems. He is also recognized as a Forbes 30 Under 30 honoree and regularly shares hands-on insights on LLMs, MLOps, and evaluation for real-world applications.

Before Deepchecks, Philip led data science research teams in the Israel Defense Forces, combining algorithmic research with applied delivery across multiple ML domains. He holds an MSc in Electrical and Electronics Engineering from Tel Aviv University, where his research proposed a new ML algorithm combining neural networks with gradient boosting and was accepted to IJCAI 2019, and a BSc in Physics from the Hebrew University of Jerusalem. His work bridges rigorous research foundations with building practical tools that help teams iterate on AI quality, reliability, and compliance.
Smiling man with short brown hair wearing a black shirt against a plain white background.
Liran Jdanov
Senior Solutions Architect
at Nebius
Read more
Liran Jdanov is a Senior Solutions Architect at Nebius, where he helps organizations design and deploy AI and machine learning workloads on large-scale cloud infrastructure.

His work focuses on enabling companies to adopt modern AI technologies through scalable cloud architectures, automation, and production-ready ML environments. Before joining Nebius, Liran led solution engineering and platform architecture initiatives for SaaS and private cloud products, guiding teams through large-scale infrastructure design, CI/CD automation, and cloud migrations. Earlier in his career, he worked on complex telecom and distributed systems projects at Ericsson and Motorola.

Liran holds a Master’s degree in Computer Engineering and has extensive experience in cloud architecture, infrastructure automation, and enterprise-scale systems.
Portrait of a smiling man with short dark hair wearing a black shirt and backpack against a light gray background.
Simon Karasik
Lead of research infrastructure
at Nebius AI R&D
Read more
Simon Karasik is a Senior Machine Learning Engineer and Team Lead at Nebius, where he leads research infrastructure for large-scale AI development. His work focuses on building systems for training and running large language models across thousands of GPUs, as well as developing infrastructure for modern AI agents and research workflows.

Before joining Nebius, Simon worked on large-scale machine learning systems at Yandex, where he developed and deployed models for high-load production environments. He holds a degree in Applied Mathematics from Belarusian State University and specializes in ML infrastructure, large-scale model training, and practical AI system design.
Portrait of a man with short dark hair and beard wearing a black shirt against a red background.
Yuval Belfer
Senior Developer Advocate
at AI21 Labs
Read more
Yuval Belfer is a Senior Developer Advocate at AI21 Labs, specializing in Large Language Models, RAG systems, and AI agents. He works at the intersection of research and real-world GenAI deployment, helping teams design, evaluate, and scale production-ready AI systems.

Yuval holds an MSc in Computer Science from the Weizmann Institute of Science, where he conducted research on neural network theory and training dynamics. In addition to his industry role, he teaches advanced courses on Generative AI and LLM development at Reichman University, Google & Reichman Tech School, and Nebius Academy. His expertise spans model behavior, evaluation frameworks, agent orchestration, and system architecture, bridging deep technical understanding with practical implementation.
Portrait of a man with a beard and mustache, wearing a dark shirt, against a light background.
Tom Braude
AI Tech Lead (Cyber & AI startup), ex-Microsoft
Read more
Tom Braude is an AI Research Engineer currently working at a stealth startup, previously Technical Lead at AI21 Labs, where he focused on planning and reasoning for agentic systems. His work spans foundation model training, fine-tuning, and the development of advanced LLM-based systems for real-world applications.

Tom previously served as an NLP Researcher at Microsoft AI, where he developed state-of-the-art models for task extraction and conversational understanding, supporting both research and production systems. He holds an MSc in Computer Science from Reichman University, specializing in multimodal learning at the intersection of computer vision and NLP, and a BSc in Computer Science from Ben-Gurion University. His expertise combines deep research in language models with practical experience in building reasoning-driven AI systems.
Smiling man with curly hair, beard, and glasses wearing a white shirt against a plain light background.
Asaf Joseph Gardin
Senior Software Engineer x Deep Learning Engineer at AI21 Labs
Read more
Asaf Gardin is a Senior Software Engineer on the Inference Team at AI21 Labs, specializing in large-scale LLM inference, performance optimization, and production deployment. He works on deploying and stabilizing advanced models including Jamba and Mamba, focusing on reliability, scalability, and real-world system performance.

Asaf is an active contributor to vLLM and has driven production deployments of LLM systems used across the organization. His work spans inference infrastructure, integration of SSMs, scheduler optimization, developing cross-language SDKs, and low-level debugging of distributed systems under memory and performance constraints. He pursued an MSc in Computer Science from Reichman University and holds a BSc in Computer Science from Holon Institute of Technology, bringing strong systems engineering expertise to modern AI infrastructure.
Elik Sror
Algorithm Group Manager at WSC Sports
Read more
Elik Sror is an AI & NLP Leader and Algorithm Group Leader at WSC Sports, where he leads multiple teams building the next generation of AI-driven content generation engines. His work focuses on LLMs, retrieval systems, and agentic AI, tackling large-scale NLP, vision, and speech challenges to deliver production-grade generative capabilities.

Alongside his industry leadership, Elik is an active lecturer and consultant, having created and taught NLP courses in academic settings and mentoring teams end-to-end — from problem definition and data pipelines to model design and integration. He holds an M.Sc. in Electrical Engineering from Tel Aviv University, specializing in computer vision and machine learning, and a B.Sc. in Electrical Engineering from Ben-Gurion University. His background combines strong research foundations with hands-on experience scaling applied AI products across multiple domains.
Yehoshua (Shuki) Cohen
 VP Data at AI21 Labs
Read more
Yehoshua (Shuki) Cohen is VP Data at AI21 Labs and an AI Evangelist focused on the intersection of AI, data, and product. He leads company-wide data strategy across data engineering, analytics, and data science, driving a data-first culture and enabling intelligent decision-making at scale.

Beyond his leadership role, Shuki is a well-known educator and community builder: he creates practical data science content through his “One Shot Learning” channel and co-founded JerusML, the Jerusalem AI community. He holds an M.Sc. (Cum Laude) in Industrial Engineering from Tel Aviv University and a B.Sc. (Summa Cum Laude) in Industrial Engineering & Management from Ben-Gurion University, combining strong analytical foundations with hands-on experience in applied machine learning, large-scale data systems, and product-driven AI adoption.
Yonatan Glassner
Senior ML engineer at NVIDIA
Yehoshua (Shuki) Cohen is VP Data at AI21 Labs and an AI Evangelist focused on the intersection of AI, data, and product. He leads company-wide data strategy across data engineering, analytics, and data science, driving a data-first culture and enabling intelligent decision-making at scale.

Beyond his leadership role, Shuki is a well-known educator and community builder: he creates practical data science content through his “One Shot Learning” channel and co-founded JerusML, the Jerusalem AI community. He holds an M.Sc. (Cum Laude) in Industrial Engineering from Tel Aviv University and a B.Sc. (Summa Cum Laude) in Industrial Engineering & Management from Ben-Gurion University, combining strong analytical foundations with hands-on experience in applied machine learning, large-scale data systems, and product-driven AI adoption.

Building AI expertise in Israel

By bringing together engineers, instructors, and industry experts across Israel, we support talent growth and help strengthen Israel's position on the global AI stage.
Join the community
Three young adults with ID lanyards smiling in an indoor setting, with two men and one woman.Smiling woman with curly hair in a beige jacket engaging in conversation with two people in an indoor setting.Two men wearing gray t-shirts and conference badges, one giving a thumbs up, smiling at the camera during a graduation event.A woman wearing glasses and a lanyard points to information on a wall poster while another person listens.Smiling woman with glasses embraces a man wearing a green shirt in an indoor setting.A man points to a board while a teenage boy looks on attentively in a hallway.
Three young adults with ID lanyards smiling in an indoor setting, with two men and one woman.Smiling woman with curly hair in a beige jacket engaging in conversation with two people in an indoor setting.Two men wearing gray t-shirts and conference badges, one giving a thumbs up, smiling at the camera during a graduation event.A woman wearing glasses and a lanyard points to information on a wall poster while another person listens.Smiling woman with glasses embraces a man wearing a green shirt in an indoor setting.A man points to a board while a teenage boy looks on attentively in a hallway.

This course is completely free of charge. It's our way of contributing to the Israeli AI community.

Both tracks require successfully completing
all admission stages

Self-paced independent learning
Unsupervised Learning Track

A great fit for independent learners who don’t require mentorship

All lectures (online & in person)

All learning materials & lecture recordings

Access to the Slack learning community

Full program
Supervised Learning Track

Full program access with mentorship, project reviews, cloud credits, and career support—designed for learners who want guided practice and a structured path to strong results

All in Unsupervised Learning Track plus:

Full individual supervision

Capstone projects with reviews & mentor feedback

Cloud credits for hands-on practice

Program certificate upon successful completion

Career support (mock interviews, HR workshops, industry prep)

Access to the Nebius global community and network

Any other questions?

How does the application process work?
What level of coding skills is required to enter the program? Do you require knowledge of specific languages?
Do I need to know machine learning to take this course?
What is the time commitment for this program? Can I combine it with work or academic studies?
Why is this program free?
What is the difference between the Supervised Learning (Guided) Track and the Unsupervised Learning (Independent) Track?
What is the language of the program?
Where does the program take place?
What is Nebius’ role in AI education?
Do I get a certificate at the end of the program?
Cookie Settings

Cookies necessary for the correct operation of the site are always enabled. Other cookies are configurable.

Always active

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Noa Lubin
VP Technology
at Israel AI Agency
Noa Lubin is VP of Technology at the Israel AI Agency and an AI leader with extensive experience in NLP, healthcare, and space technologies. Previously Director of Data Science at Fido, she has held roles at NASA, Amazon, Diagnostic Robotics, Elbit Systems, and the Israeli Aerospace Industry, leading and applying machine learning solutions in high-impact domains.

Noa holds an MSc in Computer Science (Magna Cum Laude) from Bar-Ilan University, where she specialized in Natural Language Processing under the supervision of Prof. Yoav Goldberg, and a BSc in Electrical Engineering (Summa Cum Laude) from the Technion. Alongside her industry work, she is an active lecturer and public speaker in AI and data science, combining deep technical expertise with a strong commitment to socially impactful AI initiatives.