Building an on-premise AI infrastructure is an important task that requires careful planning, investing in the right technology, and following best practices. Unlike cloud-based solutions, an on-premise setup gives you more control over your data, better security, and the ability to customize the system to meet your specific needs. However, it also requires technical knowledge, resources, and regular maintenance to work effectively.

Businesses are realizing that proving AI can work is no longer enough. To succeed, AI initiatives must deliver measurable value and remain adaptable to long-term needs. The shift from proof of concept (PoC) to proof of value (PoV) represents a fundamental change—one that emphasizes outcomes over feasibility and ensures AI solutions address real business challenges.

Perishable goods present unique challenges for supply chains. With short shelf lives, unpredictable demand, and the need for consistent cold chain management, the margin for error is slim. AI, and AI agents - new digital colleagues in supply chain teams - in particular, offers a way to address these complexities, providing advanced capabilities for improving forecasting accuracy, enhancing visibility, and building resilience. 

The energy sector is a cornerstone of modern society, powering economies and enabling daily life. As demand for electricity grows, power generation companies and utilities face unprecedented challenges, including grid reliability, environmental concerns, and the integration of renewable energy sources. These issues are compounded by the urgency of reducing carbon emissions and transitioning to sustainable energy systems.

As AI is taking a larger role in society and public discourse, and more and more people are being exposed to it, it is ever more important for everyone to understand how AI works. Understanding how Machine learning models, the most prominent type of models within AI, are trained, results in a much better grasp of the capabilities and limitations of AI. This article gives a high level explanation of how machine learning models are trained and what this means for data science projects.

AI is transforming industries, but many businesses approach it with outdated assumptions. The "build vs. buy" debate oversimplifies a complex decision. Instead of choosing between in-house development and off-the-shelf solutions, businesses should rethink their entire approach to AI - focusing on long-term adaptability, the true cost of ownership, and where they should not invest.

Organizations are increasingly turning to AI tools like Microsoft Co-Pilot to enhance productivity and streamline workflows. Designed to work seamlessly within the Microsoft 365 ecosystem, Co-Pilot enables smarter collaboration, faster data access, and automation of routine tasks. While the potential benefits are substantial, successful implementation requires careful planning to navigate challenges such as data preparation and governance.

In data science, the formulation of the problem is a critical step that significantly influences the success of any project. Properly defining the problem not only sets the direction for the entire analytical process but also shapes the choice of methodologies, data collection strategies, and ultimately, the interpretation of results. For data scientists, a well-formulated problem helps in honing in on the right questions to ask, allowing them to design experiments and models that are aligned with business objectives. It ensures that the analytical effort is relevant and impactful, leading to actionable insights rather than merely technical achievements.

Deploying AI systems is just the beginning. To create business impact and realize value, these systems must be sustained to remain reliable, adaptive, and compliant over time. AI sustainment is a strategic approach to extend the lifecycle of AI models, ensuring they are performant, scalable, and aligned with business needs. Algorithma emphasizes a proactive methodology that continuously improves models, manages data effectively, and follows responsible AI guidelines—maximizing value and maintaining a competitive edge.

AI is fundamentally changing how organizations operate, lead, and adapt. Beyond being a catalyst for increased productivity, it represents a novel form of capital that, when effectively harnessed, can reshape the operational and competitive landscape. Successful AI adoption requires more than technical expertise—it calls for rethinking team dynamics, leadership, and how organizations manage ongoing change.

The 2024 Nobel Prize in Physics highlights groundbreaking work done by John J. Hopfield and Geoffrey E. Hinton on neural networks, where they developed models like the Hopfield Network and the Boltzmann Machine, inspired by the behavior of physical systems. Their pioneering work in the 1980s laid the foundation for the machine learning revolution that took off around 2010. This award celebrates their contributions to the foundational technologies driving modern machine learning and artificial intelligence. The exponential growth in available data and computing power enabled the development of today’s artificial neural networks, often deep, multi-layered structures trained using deep learning methods. In this article we will dive into their discoveries and explain how these breakthroughs have become central in AI applications. 

To achieve sustainable cost savings, businesses must first gain a deep understanding of their cost landscape—the key cost buckets and areas of expenditure that impact overall financial performance. By mapping these costs, companies can identify where inefficiencies lie, making it easier to target specific areas for savings while improving operational performance. This approach ensures that cost-cutting efforts are strategic, sustainable, and aligned with long-term business goals. AI and advanced analytics can play a critical role in each area of the cost landscape, enabling smarter decision-making, automation, and optimization throughout the organization.

In this update, Jonathan Anderson (our CTO) explains the new DSPY framework, designed to simplify and strengthen control over large language models (LLMs). LLMs, while transformative, can be unpredictable, often behaving like “black boxes.” DSPY addresses this by offering a structured approach to interaction, reducing the need for prompt tuning and making model behavior consistent and predictable.

In this video Simon, one of our Data Scientists, takes a first look at the o1 preview model. OpenAI says the model is part of a series of reasoning models for solving hard problems. With claims that performance is on par with PhD students within fields like physics and chemistry within some benchmarks. To test it, Simon gives it a math problem, and quickly analyzes its solution. Based on this, Simon discusses potential strengths and weaknesses of the model and gives an overall impression with thoughts about the future potential of this type of model.

Machine learning (ML) is currently transforming various fields, such as healthcare, finance, and creative industries. However, as data and problems become more complex, classical computing struggles to scale ML algorithms efficiently. Key challenges include the time and computational resources needed to train models on large datasets, optimize deep learning architectures, and perform tasks like data classification and clustering. These limitations drive interest in exploring quantum computing. 

Artificial intelligence is an operational necessity in many industries, in particular financial services, driving everything from credit scoring to fraud detection. But with great power comes great responsibility: AI systems, if not managed properly, can reinforce biases and inequalities, leading to unfair lending, discriminatory insurance pricing, or biased fraud alerts. In finance, bias in AI can mean denying loans to certain groups, charging higher premiums unfairly, or disproportionately flagging transactions as suspicious—all of which have significant real-world impacts on people's lives. As AI becomes more central to finance, effective data governance is key to ensuring these systems are fair, transparent, and accountable.

Graph neural networks (GNNs) offer transformative potential for businesses by uncovering hidden patterns and relationships within complex data. From detecting fraud to optimizing supply chains and accelerating drug discovery, GNNs enable smarter decision-making and drive operational efficiency. Unlike traditional machine learning models that analyze data points in isolation, GNNs excel at identifying connections and patterns within the data. For business leaders, this technology presents an opportunity to unlock new avenues for growth and innovation, maximizing the potential of their data.

Companies are increasingly relying on machine learning models to make critical decisions. ML models come with a fundamental assumption: they expect the future to look like the past. In reality, the world is constantly changing, and so is the data it generates. This change, known as data drift, can silently undermine the performance of your models, leading to poor decisions, increased costs, and missed opportunities.

Sweden's major population centers, including Gothenburg, Stockholm, and Malmö, are faces with a looming threat of power shortages due to capacity constraints in the national grid. Property owners, transportation sectors, and heavy industries will face challenges to drive their business. AI is part of the toolbox to solve this - but getting started is key.

The ability to leverage the combined strengths of machine learning and optimization to enhance decision-making processes can significantly transform business operations. By integrating these technologies, businesses can achieve increased efficiency, reduce operational costs, and improve overall outcomes. This transformative potential is realized through practical applications in decision-making, whether by supporting human decisions or performing them autonomously.