Masterclass

GenAI timeseries forecasting for every industrial company

Build your own forecaster with GenAI. No coding, no ML expertise, no expensive licenses.

Industrial forecasting, finally within reach 

Your machines, production lines and systems generate data around the clock. Energy consumption on CNC machines, sensor readings, production volumes, sales figures, it is all there. But turning that data into reliable predictions has traditionally required a specialist team, months of development time and costly infrastructure.

GenAI has changed that equation. In this 3-hour masterclass, Sirris expert Mihail Mihaylov shows you how to build a forecaster that matches industrial benchmarks without writing a single line of code, without ML expertise and without expensive licenses. No feature engineering, no model training, no algorithm selection. The model works out of the box. 

 

For whom?

  • Maintenance managers and engineers looking to move from reactive to predictive maintenance
  • Production planners and operations managers who need reliable demand and capacity forecasts
  • Data owners and technical profiles wanting to apply GenAI without complex ML pipelines
  • Anyone who collects operational or sensor data and wants to extract more value from it
  • Professionals in energy, utilities, logistics or any data-intensive sector looking to apply forecasting without a specialist team 

 

Key takeaways

GenAI foundation models such as Chronos-bolt, TimesFM and TimeGPT can match, and in several cases outperform, the best traditional forecasting models, at a fraction of the cost and effort. Sirris has tested this against established industrial benchmarks. The results speak for themselves. 

 

After this session, you will know: 

  • How timeseries forecasting works and which accuracy metrics matter in practice
  • Which use cases are most relevant for manufacturing: predictive maintenance, energy consumption (CNC), production planning, sales forecasting and cost estimation
  • When GenAI outperforms classical ML, and when it does not
  • How to build your own forecaster in an afternoon using free, open-source models
  • Why no ML expertise is needed: the model handles feature extraction, algorithm selection and training automatically
  • How coding assistants accelerate development, and what pitfalls to avoid
  • How to get started immediately with the step-by-step recipe from Sirris  

Programme

This 3-hour masterclass is structured as follows: 

TimeSubject
12:00 – 13:00

Welcome & lunch 

  • Registration and network lunch
13:00 – 14:00

Intro: Purpose and expectations

  • Why forecasting matters and what you will gain from the session.
  • Timeseries forecasting explained: metrics like WAPE and sMAPE, univariate vs multivariate approaches.
  • Classical ML methods: Overview of techniques such as XGBoost with strengths and limitations.
  • Understanding foundation models: how they work, why they are flexible, and how they differ from LLMs. 
14:00 – 15:00

Time series foundation models 

  • Models such as Chronos-bolt, TimesFM and TimeGPT, with pros and cons.
  • Experimental results: What our tests reveal about accuracy, speed and cost compared to industrial benchmarks.
  • Coding assistants as accelerators: How Codex and Claude Code support fast , safe development and what pitfalls to avoid. 
15:00 – 16:00

The Sirris recipe 

  • A practical step-by-step guide to building your own timeseries forecaster, with concrete prompts and design steps you can use.
  • Video walkthrough: a complete demonstration of the build process in a coding assistant.
  • Live demo: Adjust parameters live and visualize the generated forecast.
  • Q&A: closing discussion 
16:00 – 16:30Networking possibility

 

About Mihail Mihaylov

Mihail Mihaylov is a GenAI expert and entrepreneur with 18 years of experience in artificial intelligence, combining a decade of work in start-ups and scale-ups with eight years of applied (gen)AI and machine-learning research.

He has led the development of AI-powered SaaS products in the energy sector and introduced new concepts and algorithms in decentralized agent-based systems. His work bridges deep technical insight with practical industrial applications, helping organisations adopt cutting-edge AI technologies with confidence. 
 

Flemish companies can take advantage of a subsidy from VLAIO for this course

This programme is part of Industriepartnershap in which 13 Flemish innovation partners offer an integrated service to stimulate growth and innovation in the Flemish industry in the 6 following themes: AI, Cybersecurity, Circularity, Digitisation, Climate & Energy and Industry 4.0. They do so under the leadership of Agoria and Sirris and with the support of Agentschap Innoveren & Ondernemen.

Banner industriepartnerschap Sirris Agoria

The material for this masterclass was developed in collaboration with the FAIR special project "Trustworthiness of Generative AI in Industrial R&D".

Date

02 April 2026 | 12:00 - 16:30

Location

Sirris Gent

Technologiepark 48
9052 Zwijnaarde
Belgium

Google maps

Price

excl. VAT: Companies based in Flanders & Sirris member: € 201 | Companies based in Flanders (non-Sirris members): € 241 | Standard price: € 1338

Language

English

Status

Closed

Experts

For all practical information, information on invoicing, location etc of this event, please contact our event manager 

adres events@sirris.be

Partners

Date

02 April 2026 | 12:00 - 16:30

Location

Sirris Gent

Technologiepark 48
9052 Zwijnaarde
Belgium

Google maps

Price

excl. VAT: Companies based in Flanders & Sirris member: € 201 | Companies based in Flanders (non-Sirris members): € 241 | Standard price: € 1338

Language

English

Status

Closed

Similar activities