Eneas Lari

Eneas Lari (born November 26, 1993) is a Greek software engineer, game developer, and technical writer. His work focuses primarily on web development, with additional interests in game design, interactive features, useful applications, and small creative software projects. He describes his practice as a continuing cycle of building, learning, experimentation, and documentation.[1]

Lari publishes through a personal website that serves as both a portfolio and an open technical notebook. Its subject matter includes software engineering, coding, web infrastructure, games, and artificial intelligence. His writing combines introductory explanation with practical guidance and is intended to record the process by which he learns and builds, rather than presenting only completed outcomes.[1]

By September 2026, his most developed public body of work was a ten-part educational series on artificial intelligence. The series moved from basic definitions and data representation to problem selection, development workflows, Python, Pandas, NumPy, and the mathematics required for introductory machine learning.[1]

Software development

Lari identifies professionally as a software engineer and game developer. His main technical focus is web development, particularly the creation of clean web experiences and practical applications. He also experiments with game design, interactive functionality, and smaller creative projects, describing his objective as taking an idea from an initial possibility to something that a person can click, play, explore, or use.[1]

This work is presented as iterative rather than fixed. Lari characterizes himself as a persistent tinkerer who is commonly positioned between learning a new subject and building another version of an existing idea. His public working principles are summarized as building, learning, and sharing. Building involves turning concepts into usable or playable forms; learning involves remaining open to changes in both thought and working method; and sharing involves documenting development rather than treating a finished result as if it had appeared fully formed.[1]

His website includes a projects area intended for software, games, experiments, and ideas that have progressed into tangible work. As of September 2026, the page contained no individually published projects. His documented public output for the period therefore consisted principally of technical articles and educational material rather than a catalog of named commercial software products.[1]

Technical writing

Lari’s website functions as an open notebook for writing about software, games, artificial intelligence, and other technical subjects under active study. The format combines individual articles with longer reading series, allowing related concepts to be followed in sequence. His subjects include programming foundations, backend development, deployment, data processing, software architecture, and the practical use of artificial intelligence.[1]

A recurring feature of the writing is its emphasis on explaining why a technology is used before concentrating on implementation details. The articles place programming concepts within wider development processes and use concrete examples to make unfamiliar terminology more accessible. This approach is visible both in his artificial-intelligence series and in his work on web infrastructure.[1]

On September 21, 2026, Lari published “Understanding Nginx: From Reverse Proxy Basics to a Secure Node.js Deployment.” The article presents Nginx as more than a conventional web server. It explains how Nginx can receive incoming traffic, handle secure HTTPS connections, deliver static files, and forward requests to backend applications. By placing those functions within a Node.js deployment, the article connects infrastructure concepts with the practical operation of modern web software.[1]

Artificial-intelligence series

In 2026, Lari developed an educational series titled “Artificial intelligence.” The series was conceived as a progressive route from introductory knowledge to the design and deployment of working AI systems. Its proposed scope begins with artificial-intelligence fundamentals, Python, data, and essential mathematics. It then extends toward machine learning, deep learning, computer vision, natural-language processing, large language models, retrieval-augmented generation, deployment, machine-learning operations, and responsible AI.[1]

Ten articles had been published in the series by September 2026. Each installment emphasized clear explanation, practical examples, and hands-on learning, with the stated purpose of helping readers move from recognizing AI terminology to building and working with AI systems.[1]

Foundational concepts

“AI, Machine Learning, Deep Learning, and Generative AI,” published on June 30, 2026, introduces distinctions among several terms that are frequently treated as interchangeable. It positions machine learning as a branch of artificial intelligence and deep learning as a more specialized form of machine learning based on multilayer neural networks. Generative AI is addressed in relation to systems that produce new material rather than only categorizing or predicting existing information.[1]

On July 16, Lari expanded the series with “How Machines Learn From Data.” The article explains training as the process through which a model examines examples and adjusts its internal calculations to reduce the difference between its predictions and known outcomes. It distinguishes features—the information used by a model—from labels, which represent the outcomes it is expected to predict. It also introduces generalization as the ability to perform usefully on cases that were not present in the training data.[1]

“Features, Labels, Inputs, and Outputs,” also published on July 16, further develops the vocabulary required to define machine-learning tasks. It uses familiar prediction problems, such as estimating house prices, anticipating subscription cancellations, and identifying defective products, to demonstrate how real-world questions are translated into model inputs and expected outputs.[1]

Selecting and developing AI projects

“What Makes a Good AI Problem?” examines whether a proposed task genuinely requires artificial intelligence. Lari argues that AI is not automatically preferable to a spreadsheet formula, explicit rule, ordinary program, or human decision. A suitable AI project generally requires a clear and measurable objective, relevant historical data, recognizable patterns, repeated decisions, and enough potential value to justify the system’s added complexity.[1]

The article differentiates prediction, classification, recommendation, and optimization problems while also considering cases in which AI should remain subordinate to human judgment. It notes that decisions involving major ethical consequences, personal responsibility, or insufficient data may not be appropriate for full automation. The underlying principle is that AI should solve a defined problem and provide a meaningful advantage over simpler alternatives, rather than being adopted merely because it is fashionable.[1]

“The Complete AI Development Pipeline: From Idea to Deployed AI System” places model training within a larger engineering process. Its scope includes defining the problem, gathering and preparing data, selecting and training a model, evaluating performance, deploying the resulting system, and maintaining it after release. By treating deployment and operation as part of the pipeline, the article presents AI development as an ongoing software-engineering discipline rather than a single modeling exercise.[1]

Programming and data foundations

Lari’s introductory programming material is designed to lower the perceived threshold for beginning AI work. “Python for AI: The Minimum You Actually Need” argues that familiarity with variables, functions, loops, and conditional statements provides enough of a basis to start learning applied artificial intelligence. The article frames additional programming knowledge as something that can be acquired alongside practical projects rather than as a prerequisite that must be mastered in its entirety beforehand.[1]

“Working With Data Using Pandas” introduces the handling of tabular datasets. It reflects the view that most AI projects begin not with an advanced model but with the more ordinary work of examining, organizing, cleaning, and transforming data. Pandas is presented as a tool for working with rows, columns, and other structured information before that information enters a machine-learning workflow.[1]

“Numbers, Arrays, and NumPy: How AI Represents Data,” published on September 9, explains how real-world information is converted into numerical structures. It distinguishes NumPy arrays from standard Python lists and introduces scalars, vectors, matrices, dimensions, and shapes. Examples include the numerical representation of students, images, text, and other information that an AI model cannot process directly in its original human-readable form.[1]

Mathematics for machine learning

“The Math You Actually Need for AI,” published on September 21, connects introductory mathematics with its role in prediction systems. Using an example based on student scores, it discusses mean, standard deviation, probability, linear relationships, functions, vectors, matrices, and prediction errors.[1]

The article explains that averages and measures of variation summarize datasets, while functions and linear relationships help describe how inputs may relate to outputs. Vectors and matrices provide structures for representing multiple values, and error measures allow a developer to compare predictions with real outcomes. Mean Absolute Error is introduced as one method of expressing the average size of prediction mistakes.[1]

Lari’s treatment emphasizes interpretation over manual calculation. Numerical libraries such as NumPy can execute the arithmetic, while the practitioner’s central task is to understand what a calculation represents, why it is being performed, and how it contributes to the behavior of a model. This applied orientation links the mathematics article to the series’ broader progression from basic concepts toward the training of real machine-learning systems.[1]

Working philosophy

Across his software and writing, Lari presents curiosity and iteration as central values. He favors making ideas tangible, remaining willing to change his understanding, and sharing intermediate lessons rather than only polished conclusions. His website describes this approach as an open notebook containing notes, experiments, projects, and concepts still in progress.[1]

Lari also connects technical development with improvement in other areas of life. Outside programming, his stated interests include physical training, practicing saxophone, and listening to music. He describes these activities through the common process of developing a skill over time, whether that means producing cleaner code, increasing physical strength, or improving a musical performance.[1]