Veliki jezički modeli (LLM) Training in Belgrade

This instructor-led, live training in Belgrade (online or onsite) is designed for senior management professionals who want to understand what LLMs are, explore their potential impact on business operations, and evaluate practical uses of AI tools such as ChatGPT, Microsoft Copilot, or Grok for real-world tasks like content creation, data summarization, and decision support.

By the end of this training, participants will be able to:

• Understand what LLMs are and how tools like ChatGPT and Copilot function.
• Use prompt techniques to get practical, reliable results from LLMs.
• Evaluate real use cases such as email drafting, summarizing documents, and productivity automation.
• Identify investment opportunities and strategic applications for AI adoption.

Ova obuka uživo, koju vodi instruktor u Belgrade (online ili uživo), namenjena je timovima za stariji menadžment koji žele da razumeju stratešku vrednost LLM-ova i preduzećkih AI alata. Polaznici će istražiti kako da integrišu ove alate u visokorazredne tokove posla, kako da pišu bolje promptove i kako da procene prilike za povećanje produktivnosti i povrata ulaganja (ROI) kroz usvajanje AI-a.

Nakon završetka ove obuke, polaznici će moći da:

• razumeju kako LLM-ovi funkcionišu i kako ih primenjuju alati poput ChatGPT-a i Copilota;
• koriste interakcije zasnovane na promptovima za automatizaciju i ubrzanje zadataka;
• primenjuju AI alate na stvarne scenarije, kao što su sastavljanje e-pošte, rezimiranje izveštaja i pregled ugovora;
• procene strateške prednosti, ograničenja i pitanja licenciranja pri usvajanju LLM-ova.

This live, instructor-led session in Belgrade, whether virtual or face-to-face, is designed for mid-level to senior AI researchers, data scientists, and developers seeking to explore, refine, and apply Meta AI’s Large Language Models within various NLP contexts.

Upon finishing this module, participants will be equipped to:

• Comprehend the architectural framework and internal workings of Meta AI’s Large Language Models.
• Prepare and tailor Meta AI LLMs for distinct operational scenarios.
• Develop and execute LLM-driven features such as text condensation, dialogue systems, and sentiment evaluation.
• Improve the efficiency and deployment strategies for large language models.

Ova obuka uživo koju vodi instruktur u Belgrade (online ili uživo na lokaciji) namenjena je stručnjacima za veštačku inteligencije srednjeg nivoa, analitičarima poslovanja i tehnološkim liderima koji žele da razumeju principe generativne veštačke inteligencije i primenu LLM modela u poslovnim okruženjima. učesnici će naučiti o transformerima, inženjerstvu upita i etičkim razmatranjima u povezivanju ovih modela u realnim rešenjima.

Na kraju ove obuke, učesnici će biti u mogućnosti da:

• Razumeju osnovne principe generativne veštačke inteligencije i modela velikih jezika.
• Implementiraju i prilagode (fine-tune) LLM modele za specifične poslovne primene.
• Primene tehnike inženjerstva upita za optimalne izlaze modela.
• Prepoznaju etička pitanja i upravlja rizicima u upotrebi LLM modela.

Ova obuka pod vođstvom instruktora u Belgrade (online ili uživo) namenjena je profesionalcima iz oblasti veštačke inteligencije srednjeg nivoa, etičarima, naučnicima za podatke i inženjerima, kao i donosiocima odluka i zainteresovanim stranama koji žele da razumeju i savladaju etički pejzaž LLM modela.

Na kraju ove obuke, učesnici će moći da:

• Prepoznaju etičke probleme i izazove povezane sa LLM modelima.
• Primene etičke okvire i principe na uvođenje LLM modela.
• Ocene društveni uticaj LLM modela i ublaže potencijalne rizike.
• Razviju strategije za odgovoran razvoj i korišćenje veštačke inteligencije.

This instructor-led, live training in Belgrade (online or onsite) targets intermediate-level NLP practitioners, data scientists, content creators, translators, and global businesses interested in using LLMs for language translation and multilingual content creation.

By the end of this training, participants will be able to:

• Understand the principles of cross-lingual learning and translation with LLMs.
• Implement LLMs for translating content between various languages.
• Create and manage multilingual datasets for training LLMs.
• Develop strategies for maintaining consistency and quality in translation.

This instructor-led live training in Belgrade (delivered online or onsite) is tailored for mid-level financial analysts, data scientists, and investment professionals looking to harness Large Language Models (LLMs) for market analysis and forecasting.

Upon completion of the program, participants will be equipped to:

• Grasp how LLMs are applied within financial market analysis.
• Leverage LLMs to interpret financial news, reports, and data to derive market insights.
• Construct predictive models targeting stock prices, market trends, and economic indicators.
• Embed LLM-derived insights into investment decision-making workflows.

This instructor-led, live training Belgrade (online or onsite) is designed for intermediate-level environmental scientists, researchers, data analysts, and policy makers or advocates interested in applying LLMs to environmental modeling and analysis.

Upon completion of this training, participants will be able to:

• Grasp the role of LLMs in environmental science.
• Apply LLMs to analyze and model environmental data.
• Evaluate LLM outputs for environmental impact assessments.
• Effectively communicate findings to influence policy and conservation initiatives.

This instructor-led, live training in Belgrade (online or onsite) targets intermediate-level VR and AR developers, game designers, and AI engineers who wish to incorporate LLMs into VR and AR applications to create more engaging and responsive environments.

By the end of this training, participants will be able to:

• Understand the role of LLMs in creating immersive VR and AR experiences.
• Develop VR and AR applications that utilize LLMs for interactive dialogues and content creation.
• Integrate LLMs with VR and AR development tools for enhanced user engagement.
• Apply best practices for designing AI-driven narratives and interactions in virtual spaces.

This instructor-led, live training in Belgrade (online or onsite) is aimed at intermediate-level data scientists, machine learning engineers, and software developers who wish to apply Large Language Models (LLMs) to multimodal data for advanced AI applications.

By the end of this training, participants will be able to:

• Understand the principles of multimodal learning with LLMs.
• Implement LLMs to process and analyze text, image, and audio data.
• Develop applications that leverage the strengths of multimodal data integration.
• Evaluate the performance of multimodal LLM systems.

Ova obuka sa trenerom uživo u <lok> (online ili na lokaciji) namenjena je sajberbebednjacima i naučnicima za podatke srednjeg nivoa koji žele da iskoriste LLM-ove za unapređenje bezbednosnih mera i uvida u pretnje.

U toku ove obuke, učesnici će steći sposobnost za:

• Razumevanje uloge LLM-ova u sajberbezbednosti.
• Primenu LLM-ova za otkrivanje i analizu pretnji.
• Korišćenje LLM-ova za automatizaciju bezbednosti i reagovanje na incident.
• Integraciju LLM-ova sa postojećom bezbednosnom infrastrukturom.

This instructor-led, live training in Belgrade (online or onsite) is designed for intermediate-level data scientists and business analysts who wish to utilize large language models (LLMs) to forecast trends and behaviors in various industries.

By the end of this training, participants will be able to:

• Understand the fundamentals of LLMs and their role in predictive analytics.
• Implement LLMs to analyze and forecast data in various industries.
• Evaluate the effectiveness of predictive models using LLMs.
• Integrate LLMs with existing data processing pipelines.

This instructor-led, live training in Belgrade (online or onsite) is designed for intermediate-level data scientists aiming to develop a comprehensive understanding and practical skills in both Large Language Models (LLMs) and Reinforcement Learning (RL).

By the conclusion of this training, participants will be able to:

• Understand the components and functionality of transformer models.
• Optimize and fine-tune LLMs for specific tasks and applications.
• Understand the core principles and methodologies of reinforcement learning.
• Learn how reinforcement learning techniques can enhance the performance of LLMs.

This instructor-led, live training in Belgrade (online or onsite) is aimed at intermediate-level content creators, marketers, and educational technologists who wish to harness the power of LLMs for generating high-quality, diverse, and engaging content across various domains.

By the end of this training, participants will be able to:

• Understand the capabilities of LLMs and their application in content generation.
• Set up and use LLMs for generating various types of content.
• Apply best practices for prompting and fine-tuning LLMs to produce desired outputs.
• Evaluate the quality of AI-generated content and refine it for specific audiences.
• Explore advanced techniques for creative and multi-modal content generation with LLMs.

This instructor-led, live training in Belgrade (online or onsite) is designed for educators, EdTech professionals, and researchers of all experience levels who aim to leverage LLMs to create personalized educational experiences.

Upon completion of this training, participants will be able to:

• Grasp the architecture and capabilities of LLMs.
• Identify opportunities for personalizing educational content using LLMs.
• Design adaptive learning platforms that employ LLMs for content personalization.
• Implement LLM-driven strategies to boost student engagement and learning outcomes.
• Evaluate the effectiveness of LLMs in educational settings and make data-driven decisions for

This instructor-led, live training in Belgrade (online or onsite) is designed for intermediate-level ML practitioners and AI developers aiming to fine-tune and deploy open-weight models such as LLaMA, Mistral, and Qwen for specific business or internal applications.

By the end of this training, participants will be able to:

• Understand the open-source LLM ecosystem and key differences between models.
• Prepare datasets and fine-tuning configurations for models like LLaMA, Mistral, and Qwen.
• Execute fine-tuning pipelines using Hugging Face Transformers and PEFT.
• Evaluate, save, and deploy fine-tuned models in secure environments.

Ova trenerom vođena, uživa trening na lokaciji Belgrade (online ili na lokaciji) namenjen je softverskim developerima i naučnicima za podatke od početnog do srednjeg nivoa znanja koji žele da implementiraju LLM-ove u sisteme za prepoznavanje i sintezu govora.

Nakon završetka ovog treninga, učesnici će biti u stanju da:

• Razumeju ulogu LLM-ova u tehnologijama govora.
• Implementiraju LLM-ove za tačno prepoznavanje govora i prirodnu sintezu govora.
• Integrišu LLM-ove sa motorima za prepoznavanje govora i sintezatorima govora.
• Procenjuju i poboljšavaju performanse sistema za govor korišćenjem LLM-ova.
• Prate trenutne trendove i buduća pravca u tehnologijama govora.

This instructor-led, live training in Belgrade (online or onsite) targets beginner to intermediate customer support and IT professionals who wish to implement LLMs to create responsive and intelligent customer support chatbots.

By the end of this training, participants will be able to:

• Understand the fundamentals and architecture of Large Language Models (LLMs).
• Design and integrate LLMs into customer support systems.
• Enhance the responsiveness and user experience of chatbots.
• Address ethical considerations and ensure compliance with industry standards.
• Deploy and maintain an LLM-based chatbot for real-world applications.

Ova obuka sa instruktorom uživo u Belgrade (online ili na lokaciji) namenjena je data scientistima i AI inženjerima srednjeg nivoa koji žele da prilagode velike jezičke modele povoljnije i efikasnije koristeći metode poput LoRA, Adapter Tuning-a i Prefix Tuning-a.

Na kraju ove obuke, učesnici će moći da:

• Razumeju teoriju iza pristupa efikasnog prilagođavanja parametara.
• Implementiraju LoRA, Adapter Tuning i Prefix Tuning koristeći Hugging Face PEFT.
• Uporedne perfomaanse i troškove PEFT metoda u odnosu na kompletno prilagođavanje.
• Implantiraju i skaliraju prilagođene LLM-ove sa smanjenim zahtevima za izračunavanje i skladištenjem.

This instructor-led, live training in Belgrade (online or onsite) targets intermediate to advanced machine learning engineers, AI developers, and data scientists seeking to learn how to utilize QLoRA for the efficient fine-tuning of large models tailored to specific tasks.

By the conclusion of this training, participants will be capable of:

• Understanding the theory behind QLoRA and quantization techniques for LLMs.
• Implementing QLoRA to fine-tune large language models for domain-specific applications.
• Optimizing fine-tuning performance on limited computational resources using quantization.
• Deploying and evaluating fine-tuned models efficiently in real-world applications.

This instructor-led, live training in Belgrade (online or onsite) targets intermediate-level data and marketing professionals who wish to use LLMs to analyze and interpret public sentiment from various text sources such as social media posts, product reviews, and customer feedback.

By the end of this training, participants will be able to:

• Understand the principles of sentiment analysis and its application using LLMs.
• Preprocess and prepare datasets for sentiment analysis.
• Train and fine-tune LLMs to accurately reflect sentiment in text.
• Analyze sentiment in real-time from social media and other text sources.
• Integrate sentiment analysis findings into business strategies and decision-making processes.

This instructor-led, live training in Belgrade (online or onsite) is designed for intermediate-level software developers and technical writers who aim to leverage LLMs to streamline their coding workflow and create detailed, comprehensive documentation.

Upon completing this training, participants will be able to:

• Grasp how LLMs contribute to automating code generation and software documentation.
• Apply LLMs to generate precise and efficient code snippets and documentation.
• Incorporate LLMs into the software development lifecycle to boost productivity.
• Uphold high-quality documentation standards using automated solutions.
• Navigate ethical considerations and adopt best practices for utilizing AI in software development.

Ova obuka koju vodi instruktor u <lokalnoj> sredini (online ili na lokaciji) namenjena je poslovnim profesionalcima i analitičarima podataka srednjeg nivoa koji žele da iskoriste snagu LLM za ekstrakciju poslovnih uvida.

Nakon završetka ove obuke, učesnici će moći da:

• razumeju osnove i primene LLM u kontekstu poslovne analitike;
• primene LLM za analizu velikih skupova podataka i ekstrakciju značajnih uvida;
• integrišu analitiku vođenu LLM u procese strateškog poslovnog odlučivanja;
• procene etička pitanja i najbolje prakse za korišćenje LLM u poslovanju;
• predvide buduće trendove veštačke inteligencije i pripreme se za evoluirajući pejzaž poslovne analitike.

This live, instructor-led training in Belgrade (online or on-site) is intended for developers and data scientists at intermediate to advanced levels who seek to master LlamaIndex for creating innovative applications powered by LLMs.

By the end of this training, participants will be able to:

• Install and configure LlamaIndex for LLM integration.
• Index and query custom datasets using LlamaIndex to enhance LLM capabilities.
• Design and develop sophisticated applications leveraging LlamaIndex and LLMs.
• Understand and apply best practices for working with LLMs and LlamaIndex.
• Navigate the ethical considerations involved in deploying LLM-powered applications.

This instructor-led, live training in Belgrade (online or onsite) targets intermediate-level AI researchers, machine learning professionals, and data scientists who wish to use LlamaIndex to enhance AI model capabilities, making them more accurate and reliable for various applications.

By the end of this training, participants will be able to:

• Understand the principles and components of LlamaIndex.
• Ingest and structure data for use with LLMs.
• Implement context augmentation to improve AI model performance.
• Integrate LlamaIndex into existing AI systems and workflows.

Ovo instruktorsko, uživo obučavanje u Belgrade (online ili na lokaciji) namenjeno je profesionalima srednjeg nivoa koji žele da iskoriste snagu inženjeringa promptova i učenja u nekoliko koraka da optimizuju performanse LLM modela za primenu u stvarnom svetu.

Do kraja ovog treninga, učesnici će moći da:

• Razumeju principe inženjeringa promptova i učenja u nekoliko koraka.
• Dizajniraju efikasne promptove za različite NLP zadatke.
• Iskoriste tehnike few-shot prilagodbom LLM modela uz minimalne količine podataka.
• Optimizuju performanse LLM modela za praktične primene.

Delivered as an instructor-led, live session—available either online or on-site—this course is tailored for senior engineers, AI specialists, and localization leaders aiming to implement large language model (LLM) solutions for automated translation, quality assessment, and enterprise-grade governance.

Upon completion, participants will have the capability to:

• Construct enterprise-grade LLM localization pipelines that integrate both open-source and proprietary models.
• Implement automated QA workflows and quality metrics to ensure translation consistency.
• Establish robust governance and approval frameworks for multilingual content production.
• Deploy scalable, auditable LLM-based localization systems within secure environments.

This live, instructor-led training held in Belgrade empowers data engineers with the skills necessary to integrate LLMs into SQL environments. Participants will learn to construct natural language pipelines, apply AI-driven optimization techniques, and design secure enterprise workflows for automated data analysis.

This instructor-led training in Belgrade empowers professionals to create LangGraph-based finance solutions. Learners will master implementing workflows that align with regulatory requirements, integrating data standards, and deploying systems with robust governance, observability, and compliance controls.

This live training in Belgrade empowers practitioners to design, construct, and optimize multimodal LLM workflows using Vertex AI. The curriculum covers Gemini models, long-context reasoning, and the creation of pipelines that integrate text, audio, and image data. Through hands-on labs, participants gain practical skills essential for real-world deployment.

Discover the application of LangGraph in healthcare through this live training held in Belgrade, designed for intermediate to advanced professionals. Learn to engineer stateful, compliant workflows, integrate standards like FHIR and SNOMED CT, and navigate ethical and operational challenges within regulated environments.

This live, instructor-led training session, available either online or on-site, targets intermediate-level AI developers and localization engineers seeking to design scalable, automated translation pipelines using both proprietary and open-source LLMs.

Upon completing this training, participants will be capable of:

• Architecting and implementing translation workflows using modern LLM frameworks and APIs.
• Integrating open-source and commercial models into scalable translation systems.
• Enhancing translation quality through fine-tuning, prompt engineering, and automation.
• Implementing cost-efficient and compliant translation infrastructure for enterprise environments.

This instructor-led, live training in Belgrade is tailored for advanced AI platform engineers, AI DevOps specialists, and ML architects aiming to optimize, debug, monitor, and operate production-grade LangGraph systems, which build stateful, multi-actor LLM applications as composable graphs.

This instructor-led live training in Belgrade (online or onsite) targets intermediate-level developers who wish to learn how to use generative AI with LLMs for various tasks and domains.

By the end of this training, participants will be able to:

• Explain what generative AI is and how it works.
• Describe the transformer architecture that powers LLMs.
• Use empirical scaling laws to optimize LLMs for different tasks and constraints.
• Apply state-of-the-art tools and methods to train, fine-tune, and deploy LLMs.
• Discuss the opportunities and risks of generative AI for society and business.

Achieve mastery in AI integration with PostgreSQL through this instructor-led live training in Belgrade. Acquire the skills to configure extensions, deploy pgvector embeddings, and link LLMs for instant insights. Enhance query performance and construct intelligent systems within a hands-on live-lab setting.

Ova obuka koju vodi instruktur u Belgrade (online ili uživo na licu mesta) namenjena je AI developerima, arhitektama i produkt menadžerima srednjeg do naprednog nivoa znanja, koji žele da identifikuju i ublaže rizike povezane sa aplikacijama pogonjenim LLM-ovima, uključujući ubacivanje promptova (prompt injection), curenje podataka i nefiltrirane izlaze, uz uvođenje bezbednosnih kontrole poput validacije ulaznih podataka, nadzora sa ljudskim učešćem (human-in-the-loop) i mehanizama za zaštitu izlaznih podataka.

Na kraju ove obuke, učesnici će moći da:

• razumeju osnovne ranjivosti sistema zasnovanih na LLM-ovima.
• primene principe bezbednog dizajna na arhitekturu LLM aplikacija.
• koriste alate poput Guardrails AI i LangChain za validaciju, filtriranje i bezbednost.
• integrišu tehnike poput sandboxing-a, red teaminga i pregleda sa ljudskim učešćem u produkcijske pipeline-ove.

This instructor-led, live training in Belgrade (available online or onsite) is designed for intermediate to advanced-level professionals who wish to customize pre-trained models for specific tasks and datasets.

By the end of this training, participants will be able to:

• Understand the principles of fine-tuning and its applications.
• Prepare datasets for fine-tuning pre-trained models.
• Fine-tune large language models (LLMs) for NLP tasks.
• Optimize model performance and address common challenges.

This instructor-led, live training in Belgrade (online or onsite) is aimed at intermediate-level data scientists, AI developers, and AI enthusiasts who wish to use LLMs to perform various NLP tasks and create novel and diverse content for different purposes.

By the end of this training, participants will be able to:

• Establish a development environment with LLMs and essential tools.
• Expertly perform NLU and NLI tasks with LLMs.
• Extract, infer, and utilize knowledge graphs effectively.
• Generate and manage dialogues using LLMs for conversational applications.
• Evaluate content quality and diversity generated by LLMs and generative AI.
• Apply ethical principles, ensuring fairness and responsible use of LLMs.

This instructor-led LangGraph Fundamentals course in Belgrade empowers developers to construct graph-based LLM applications. You will acquire proficiency in core concepts such as nodes, edges, state management, and branching logic. The curriculum guides you through integrating tools, retrieval systems, and external APIs, culminating in hands-on labs for testing and deploying reliable workflows.

This instructor-led, live training in Belgrade (online or onsite) is aimed at beginner-level to intermediate-level developers who wish to use Large Language Models for various natural language tasks.

By the end of this training, participants will be able to:

• Set up a development environment that includes a popular LLM.
• Create a basic LLM and fine-tune it on a custom dataset.
• Use LLMs for different natural language tasks such as text summarization, question answering, text generation, and more.
• Debug and evaluate LLMs using tools such as TensorBoard, PyTorch Lightning, and Hugging Face Datasets.

This live, instructor-led training in Belgrade (available online or on-site) is designed for observability and SRE engineers seeking to integrate LLMs and AI into their monitoring, alerting, and incident analysis workflows.

This instructor-led live training in Belgrade (online or onsite) is designed for IT support professionals who want to use AI tools and automation platforms to streamline support processes, standardize responses, and reduce manual documentation effort.

This instructor-led, live training in Belgrade (online or onsite) is aimed at intermediate-level engineers and architects who wish to use Tencent Hunyuan to deploy large and MoE models with lower latency, stronger throughput, and better cost control.

By the end of this training, participants will be able to: explain Hunyuan production deployment patterns, optimize inference performance, implement batching and quantization strategies, and plan scalable serving operations.

This instructor-led, live training in Belgrade (online or onsite) is designed for intermediate-level developers, technical product teams, and AI practitioners who wish to utilize Hunyuan models to develop multimodal applications for generating and delivering images, 3D assets, and video.

By the end of this training, participants will be able to build prompt-based workflows, generate and evaluate multimodal assets, serve outputs through applications or APIs, and integrate Hunyuan capabilities into enterprise product stacks.

Course goals focus on turning software and AI engineers into legal engineers—professionals who can effectively build AI for legal work, including eDiscovery, review, and investigations.

In two days, participants construct the full core stack: ingesting and extracting messy real-world legal data, searching and retrieving it, adding retrieval-augmented generation (RAG) with citations, ensuring privacy through local models while proving integrity, running a defensible AI review with court-ready metrics, and packaging the result for deployment.

This instructor-led, live training in Belgrade (online or onsite) is aimed at beginner-level, intermediate-level, or advanced-level AI professionals who wish to use MCP to connect AI assistants with external tools, data, and enterprise services.

By the end of this training, participants will be able to: explain MCP concepts, identify key architecture components, set up a basic integration, and apply security good practices.

This instructor-led, live training in Belgrade (online or onsite) is aimed at intermediate-level enterprise architects who wish to use Model Context Protocol to design secure, scalable, and governable agent integration platforms for enterprise environments.

By the end of this training, participants will be able to: explain MCP architecture and enterprise patterns, design secure integration platforms, apply governance and access controls, and evaluate deployment and scaling options.

Ovaj trenerom vođen, uživo obuku u Belgrade (online ili na lokaciji) namenjena je developerima, arhitektama i inženjerima platforma srednjeg nivoa znanja koji žele da koriste MCP za izgradnju pouzdanih servera i klijenata za preduzećno postavljanje i operacije.

Na kraju ove obuke, učesnici će moći da: objasne MCP arhitekturu u praksi, izgrade produktivne integracije, postave i posmatraju MCP servise, i primene obrasce verzionisanja, otpornosti i podrške.

This instructor-led, live training in Belgrade (online or onsite) is aimed at intermediate-level IT leaders, compliance professionals, security teams, and enterprise architects who wish to use sovereign AI principles and governance practices to design AI environments that protect sensitive data, support localization requirements, and reduce vendor lock-in.

By the end of this training, participants will be able to: explain sovereign AI concepts, evaluate hosting and governance options, define controls for prompts and logs, and create a practical adoption roadmap.