SIAM Company has announced the launch of the Siam ML Hack hackathon, dedicated to the application of machine learning in the oil and gas industry. The event will take place from February 28 to March 19 and will serve as a platform for developers, analysts, and machine learning specialists looking to propose innovative solutions for real-world industry challenges.
The hackathon tasks are of moderate complexity, allowing specialists of varying experience levels to test their skills. Additionally, Sergey Vadimovich Isupov, Head of the Automation Department at Siam, shared insights into the upcoming event.
In the photo: Isupov Sergey Vadimovich, Head of the Automation Department at SIAM
– SIAM Company has been operating in the oilfield services industry since 1990. How have data analysis approaches evolved over this time, and which technologies are currently in the highest demand?
Over the past decades, data collection and analysis methods in oilfield services have undergone tremendous changes. Before the 1990s, mechanical instruments were primarily used at oilfields, recording data on paper—similar to hospital ECG printouts. Analyzing such data required manual processing using rulers, compasses, and other "antique" tools. These paper tapes were collected, processed, and only then used for decision-making.
With the rise of digital technologies, data storage became electronic, enabling automated calculations and database integration. This milestone marked the beginning of our company's journey—we started developing devices that replaced outdated mechanical instruments and introduced some of the first analytical software tools.
As computing power increased and data volumes grew, more complex numerical analysis methods gained popularity. Today, we are integrating AI-powered tools that expand on traditional analytical and numerical approaches. These modern solutions operate more autonomously, processing large data streams without the need for constant manual supervision by engineers.
– How is machine learning applied in hydrodynamic studies? What types of data do you work with?
We utilize various neural network architectures, including fully connected (FC), convolutional (CNN), and recurrent (RNN) networks. Each solves specific atomic tasks, which are then combined into complex systems to tackle more advanced engineering challenges. The main types of data we work with include:
• Time series: Continuous measurements of well parameters.
• Graphical data: Images of graphs containing visual patterns.
• Tabular prior information: Reference data and various object parameters.
Recently, we have been actively researching large language models (LLMs) and developing specialized agents based on them. These agents help automate routine analytics and data processing tasks.
– How do you see the development of machine learning in oilfield services? What tasks could be automated in the future?
The trends are similar to other industries: any routine, low-intelligence tasks will gradually be automated using AI. This primarily includes data collection, classification, preprocessing, and simple interpretations. We see large language models (LLMs) and intelligent agents being used not only for analytics but also in dispatching and logistics. Ideally, humans will continue solving strategic and complex scientific (rocket science) problems, while AI will handle day-to-day operations. Of course, reaching this point will take time and effort, but early adopters will gain a significant competitive advantage.
– How do IT departments collaborate with other employees at your company?
We maintain a friendly yet productive atmosphere. Formal processes, such as requirement gathering and task tracking, go through standard tools like bug trackers and task trackers. However, we are not bound by strict regulations—if an urgent problem or new solution arises, employees can directly interact with developers or engineers.
There are cases where a user requests a quick fix, or a developer asks for real-world testing of a new feature. Sometimes, these experimental ideas evolve into full-fledged products. We strive to balance formal structure with a creative, dynamic approach.
– Will the final solutions from the hackathon be used in the company’s operations? Is further collaboration possible?
If the hackathon produces better results than our internal developments, we will immediately integrate them into our work. We are currently in an active development phase, meaning new solutions can be quickly implemented after necessary testing.
As for collaboration, this is one of the main reasons we are hosting this hackathon. We need specialists, and if we see strong teams or standout individuals, we will be happy to offer them opportunities, including full-time positions.
– Are there job opportunities for IT specialists at Siam? There are no listings on hh.ru. Does the company plan to hire IT professionals?
We don’t rely heavily on hh.ru, as finding qualified and motivated specialists there can be difficult. Instead, we are exploring alternative recruitment methods, with hackathons being one of them. We need IT professionals and are always open to proposals from potential candidates.
– What skills do IT specialists need to work in oilfield services? What is important beyond machine learning?
We look for engineers who can effectively solve problems and deliver functional products. While strong theoretical knowledge is valuable, we prioritize practical problem-solving and the ability to see projects through to completion.
We are not looking for “rock stars” or those who chase new technologies just for the sake of learning them. Instead, we prioritize reliability, engineering thinking, and a willingness to understand oilfield service specifics.
If someone is passionate about solving problems, can stay focused, and demonstrates results, they are a potential fit for our company.
– Is remote work an option for IT specialists at your company?
Yes, we adopted a distributed model long ago. Our development team members are located in at least four different cities. Remote work has proven effective, so we are comfortable with geographically dispersed teams.
– What real business challenges are behind the hackathon tracks? Can they be considered similar to the cases that SIAM specialists encounter?
The hackathon tasks are not just theoretical exercises; they are real business cases that our engineers face in their daily work. For example, the track on identifying binary features helps refine complex interpretation algorithms and improve the accuracy of final models.
Another task involves determining relevant data areas in a continuous stream of well measurements. It is crucial to automatically identify which time intervals or segments deserve further analysis and which can be discarded as irrelevant. Simply put: the first task helps interpret data, while the second one determines which data should be interpreted in the first place.
– Why were these tracks chosen? Are they aimed at innovation, new solutions, or attracting new employees to the company?
We have two main goals associated with these tasks:
Improving or finding alternative solutions:
• The second task has not yet been implemented within the company, so we hope to receive either a prototype or a fully developed solution.
• Вторая задача пока не имела готовой реализации внутри компании, поэтому мы надеемся получить либо прототип, либо полноценное решение.
Team expansion:
• The hackathon is a great opportunity to find people interested in our field. We have many ideas and projects and are constantly looking for new developers.
– What level are the hackathon tasks designed for? Who is the target audience for the event?
By nature, these tasks are of medium difficulty, possibly even closer to entry-level. The key operations involved are classification, regression, and basic data processing. However, this does not diminish their relevance and value, as such "typical" tasks make up a significant portion of real-world machine learning applications in oilfield services.
In the photo: the Siam Well Test software suite
– What led to the decision to organize a hackathon? This is the company's first such event, correct?
Yes, this is indeed our first experience hosting such an event. The idea came after we started actively recruiting young specialists with prior hackathon experience. They shared how effective this format is for fostering innovation, improving skills, and networking. We see several key goals for ourselves:
1. Increasing the company’s visibility in the information space.
2. Obtaining real solutions for real-world cases.
3. Attracting new employees—we have many plans and projects, and we want to find people willing to work on them.
– Can you share some of the most successful projects implemented with IT specialists at SIAM? What results did they bring, and how did they impact the company’s development?
SiamWellTest
This is our flagship product for interpreting hydrodynamic and gas dynamic studies. It effectively replaces imported counterparts and meets the majority of industry needs. We have successfully promoted this product in the Russian market (even though it is a niche area) and are actively developing its AI-powered automation features.
SiamEngy
A mobile application designed for a broader audience—from beginner oil workers and students to practicing engineers. It provides reference data and tools that simplify daily tasks. Currently, around 15,000 users utilize the app across both platforms (Android and iOS). For its niche, this is a very strong performance, and the app is distributed for free. We have not found any market alternatives that match it in quality and content.
SiamDataSpectrum
A new software solution for systematically storing large datasets collected from sensors. In oilfield services, such data consists of continuous, high-frequency streams—potentially terabytes of information per month—that must be stored, quickly indexed, and segmented for instant access to relevant intervals. The product is currently in beta testing within the company, and we are preparing it for market launch.
Registration for the hackathon is open until February 27 at: https://codenrock.com/contests/siam-ml-hack/
The Siam ML Hack meetup successfully took place on February 21 at 19:00 Moscow time. During the livestream, organizers provided a detailed overview of the hackathon tasks and key logistical details. If you missed the event, we recommend watching the recording!
The Siam ML Hack itself will take place from February 28 to March 19. Winners will receive valuable prizes from partners and the opportunity to collaborate with one of the leading companies in the oilfield services industry.
If you have any question connected with SIAM Company,
leave your email address and we will contact you
