Iuliia Gorshkova

When we talk about compliance, we are really talking about a massive, interconnected web of constraints. A regulation passed in Brussels might reference a standard set by ISO, which in turn modifies how a specific financial transaction is logged in a New York database. A privacy law in California might conflict slightly with a data [...]

There's a peculiar tension that lives inside every engineer who has ever shipped a system powered by machine learning. We spend months curating datasets, tweaking hyperparameters, and wrestling with loss functions until the model performs beautifully on the validation set. It feels like magic, and in many ways, it is. But then comes the moment [...]

There's a particular kind of frustration that settles in when you watch a large language model try to plan a multi-step task. It's a feeling akin to watching someone try to assemble a complex piece of furniture by reading all the instructions at once, simultaneously. The model might generate a reasonably coherent sequence of actions, [...]

There’s a particular kind of frustration that comes from watching a large language model tackle a problem it *almost* solves. It’s the feeling of seeing a brilliant student ace every practice question but completely flub the final exam because the final exam requires connecting three different chapters of the textbook, and the student can only [...]

The last time I felt a genuine sense of awe watching a software demo was about two years ago. It was a video showcasing a swarm of drones navigating a dense forest at speed, finding gaps between branches, and adjusting their flight paths in real-time. There was no central brain dictating every movement. Each drone [...]

The term agentic AI has recently exploded into the tech lexicon, often wrapped in marketing hype that promises fully autonomous digital employees. As someone who has spent years building and debugging complex systems, I find the reality far more fascinating—and fragile—than the glossy press releases suggest. To understand what an AI agent actually is, we [...]

There’s a peculiar comfort in the trajectory of the last decade of artificial intelligence. If you squint at the loss curves, the scaling laws appear almost geological in their inevitability: more parameters, more data, more compute, and the model simply gets smarter. It’s a seductive narrative because it reduces the chaotic complexity of intelligence to [...]

When we talk about intelligence, whether biological or artificial, memory isn't just a passive storage bin. It is the dynamic scaffolding upon which reasoning is built. In the early days of large language models, the prevailing assumption was that more parameters equaled better recall. We treated the model weights as a static, frozen library of [...]

Among the myriad challenges used to evaluate artificial intelligence, few are as deceptively simple as the Towers of Hanoi. With its three pegs and a stack of disks, it appears to be a straightforward exercise in recursive logic, a puzzle that a computer science undergraduate solves in an afternoon. Yet, in the context of AI [...]

When I first started building language models, I treated benchmark scores like a holy grail. If a model achieved 85% accuracy on SQuAD or 90% on GLUE, I assumed it was "smarter" than a model scoring 80%. It’s a natural assumption—metrics are supposed to be objective, right? But after years of shipping models into production [...]

When engineers talk about building AI systems, the terms "safety" and "alignment" are often used interchangeably. They appear in the same meeting notes, slide decks, and product requirements documents, usually as a single bullet point: "Ensure safety and alignment." This conflation is a category error that leads to blind spots in system design. While they [...]

When we talk about "ethical AI," the conversation often drifts toward philosophy and abstract principles. We discuss fairness, justice, and the moral implications of algorithmic decisions. While these discussions are vital, they frequently miss a critical point: AI ethics is fundamentally an engineering challenge. It is not enough to declare that an AI system should [...]

When you first see a neural network correctly classify a medical image or flag a fraudulent transaction, the immediate reaction is often a mix of awe and acceptance. The model works, so we trust it. But in high-stakes environments—like a courtroom, a surgical theater, or a financial trading floor—performance metrics alone are insufficient. The question [...]

When a large language model confidently states that Barack Obama won the Nobel Prize in Chemistry, it’s not lying. It’s not being malicious, and it’s certainly not "misunderstanding" the world in the human sense. It is, however, executing its core function with mathematical precision in a way that diverges from reality. This divergence—commonly termed a [...]

For years, the world of Artificial Intelligence has felt like a tug-of-war between two fundamentally different philosophies. On one side, you have the connectionist approach—neural networks, deep learning, the "black box" models that learn patterns from vast oceans of data. These systems, particularly Large Language Models (LLMs), are incredibly fluent, creative, and capable of astonishing [...]

There's a peculiar gravity well in modern AI discourse that pulls every conversation toward large language models. It’s understandable, of course. The sheer fluency of systems like GPT-4 is a siren song for anyone who has ever dreamed of natural communication with machines. Yet, in this rush toward statistical approximation, we seem to have collectively [...]

Every AI engineer has faced that sinking feeling. You’re reviewing a large language model’s output for a sensitive application—perhaps a medical diagnosis support system or a financial compliance checker—and you spot it. The model has confidently stated something that is factually incorrect, contextually inappropriate, or simply nonsensical. It’s not a bug in the traditional sense; [...]

When I first encountered the terms schema, ontology, and knowledge graph in the context of data engineering, I treated them largely as synonyms. It was a mistake born of enthusiasm and a lack of rigorous distinction. In the early days of a project, when the architecture is just a sketch on a whiteboard, these concepts [...]

When we build AI systems, especially those that need to reason about the world, we often stumble into a problem that seems simple at first but quickly spirals into complexity: how do we represent knowledge in a way that a machine can actually understand? Not just pattern-match, but truly comprehend the relationships between entities? This [...]

When we first started building retrieval-augmented generation systems, the process felt almost magical. We took a massive pile of unstructured text, chopped it into manageable chunks, and threw them into a vector database. A user asks a question, we embed the query, find the nearest text chunks in high-dimensional space, and feed those to an [...]

Most developers I talk to have reached a similar point of frustration. You feed a large language model a few documents, maybe a dense PDF or a chunk of internal wiki text, and ask it a specific question. The model responds with absolute confidence, citing details that sound plausible but are subtly wrong, or it [...]

When you first encounter the hype surrounding large language models, the narrative almost always revolves around the size of the context window. It’s presented as the ultimate metric of capability—the longer the window, the smarter the model. We’ve seen the numbers skyrocket from a few thousand tokens to over a million in a single generation. [...]

There's a peculiar comfort in watching a large language model lay out its thoughts step-by-step. You ask it to solve a logic puzzle, and it responds not just with an answer, but with a narrative: "First, I will identify the constraints. Then, I will map the variables. Finally, I will test the hypothesis." It feels [...]

If you've spent any significant time wrestling with large language models, you've likely hit the wall of their finite context windows. You craft a meticulously detailed prompt, feed in a long conversation history, and watch as the model slowly forgets the instructions given at the very beginning. It’s a frustrating limitation of the transformer architecture: [...]

When we talk about artificial intelligence today, the conversation almost invariably circles back to Large Language Models. These systems have moved from academic curiosity to a foundational layer of modern software, yet for many developers and engineers, they remain a kind of "black box." We feed them text, and text comes out—sometimes brilliant, sometimes nonsensical. [...]

Imagine walking into your favorite grocery store and being greeted by a robot that not only recognizes you but also remembers your usual purchases, dietary restrictions, and even your preferred brands. In-store robots equipped with advanced preference-memory capabilities are no longer just a futuristic concept; they are rapidly becoming a tangible reality in the evolving [...]

Open-source initiatives are the backbone of contemporary scientific, technological, and creative progress. They democratize access to cutting-edge tools and foster collaboration across disciplines and continents. This round-up explores some of the most influential and promising open-source projects, libraries, and datasets in various domains—including artificial intelligence, data science, web development, and more. Each entry includes a [...]