Skip to content
View Lexoniarus's full-sized avatar
😶‍🌫️
😶‍🌫️

Block or report Lexoniarus

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Lexoniarus/README.md

Alex · Lexoniarus Lab

Software Engineering · AI Engineering · Business × Data × Systems

I explore how technology, real-world processes and playful ideas can become useful, reliable software.

Portfolio & project stories · About me · Get in touch

About me

I'm Alex (Alexander Altmann), based in Berlin. I combine experience in logistics, sales controlling, ERP/CRM processes and business intelligence with hands-on software development.

I completed a Software Engineering training program, specializing in AI Engineering, at Masterschool in September 2026. My technical focus is Python, backend applications, data models, APIs, local desktop software and practical AI integration.

Some projects start with a process that could be simpler. Others start with a question such as: What if a game used real roads? Or: How much personality can a desktop character have without relying on a language model?

Selected projects

The four projects below also have detailed stories, screenshots and honest development status reports on Lexoniarus Lab.

01 · BENNO — AI-assisted B2B visit reports

Completed Masterschool MVP · Demo, not a production CRM integration

Free-form visit notes become structured report proposals. Gemini interprets the input, while deterministic application logic validates fields, asks for clarification and requires human approval before saving to a mock CRM/ERP system.

Tech: Python · Flask · SQLAlchemy/SQLite · Gemini · speech input/output prototype

Source code · Project story

02 · Aiko Desktop Pet — a character that lives on the desktop

Active Windows prototype · Local-first

An animated companion with five needs, friendship, hand-written German dialogue, mini-games and reactions to interactions. Limited local system signals, including resource usage and idle time, give Aiko context through deterministic logic. A broader assistant and understanding of media content are future goals, not current capabilities.

Tech: Python · Tkinter · modular OOP · local persistence · system-signal providers

Source code · Project story

03 · World Freight Idle — multiplayer logistics on real roads

Local multiplayer prototype · Public 24/7 operation still a goal

An idle logistics simulation with real-world road maps, routed vehicles, separate player accounts, visible shared traffic and progress while players are away. Its current main-branch baseline uses SQLite; PostgreSQL/Supabase integration is being explored separately.

Tech: Python · FastAPI · SQLite · MapLibre/OpenStreetMap · routing

Source code · Project story

04 · ComfyReview — learning from image-generation workflows

Active local application · Experimental game-related POCs

Review, rank and compare ComfyUI-generated images while preserving connections to prompts, parameters and workflows. A versioned prompt catalog and generator support repeatable experiments. The related Character Chronicles card-battler work remains a proof of concept, not a finished playable game.

Tech: Python · FastAPI · SQLite · ComfyUI · prompt/workflow metadata · local model integration

Source code · Project story

Engineering approach

  • Understand the domain first: people, workflows, data quality and failure modes matter as much as technology choices.
  • Keep responsibilities explicit: separate business rules, state, providers, persistence and UI.
  • Use AI selectively: language models can interpret ambiguous input; deterministic code should own validation and business decisions.
  • Make change safer: modular design, reproducible behavior, automated tests and clear implementation-status documentation.
  • Build to learn: prototypes are useful when their limitations are as visible as their working features.

My current toolbox includes Python, FastAPI, Flask, Tkinter, SQL, SQLite, MySQL, JavaScript/TypeScript, Astro, GitHub Actions, Docker, ComfyUI and LLM APIs. Different projects use different subsets; this is not a claim that every tool is used in every application.

Background & opportunities

Before focusing on software development, I worked in operational logistics and in sales controlling and process development. At S:FLEX (2023–2025), I independently handled many SQL queries, CRM views and data requests, took ownership of ERP/CRM user documentation, and collaborated with colleagues on process improvements and MySQL/Grafana-based business intelligence.

I'm open to professional opportunities where software and AI engineering meet business processes, data, system integration or automation.

For the full story, current project screenshots and contact options, visit lexoniaruslab.de.

Pinned Loading

  1. aiko-desktop-pet aiko-desktop-pet Public

    Local-first Windows desktop pet with Tamagotchi care, adaptive workload reactions and hand-written German dialogue.

    Python

  2. world-freight-idle world-freight-idle Public

    UI-first multiplayer logistics idle game with FastAPI, SQLite and MapLibre

    Python