Selected work across AI, HPC, and infrastructure - built where nothing existing fit, adapted where something already did.
UuedViljapuuaiad OÜ partnered with CodeZero Group on these groundbreaking projects, with CodeZero leading the technical implementation under the visionary guidance of Andy Ramgobin, the brilliant mind behind these innovative solutions.
Let's talk about what's actually needed - and whether the right answer is something new or something we've already built.
CONTACT USSee what we are shipping right now - Work Feed
A proprietary platform automating repetitive data science tasks. Streamlines data ingestion, processing, distribution, and visualization without requiring extensive coding knowledge.
A modern corporate website with elegant dark theme design, featuring responsive service cards and gradient UI elements - built as UV's own site, doubling as a working example of what the CMS underneath it can do.
A distributed compute orchestration platform for machine learning workloads. Registers and monitors GPU/CPU compute nodes, schedules job submissions against priority, fair-share and backfill algorithms, and tracks workloads through their full lifecycle. Four role-specific dashboards - System Administrator, ML Researcher, Resource Manager and Operations Monitor - sit on a JWT-authenticated REST API, backed by Delta Prism, a UV-built property graph database. Built in a one-week push in June 2025 and untouched since; currently dormant, not under active development.
and 1 more
A browser-based 3D viewer for Building Information Modelling (BIM) files in the IFC format, built on Three.js and web-ifc. Loads and switches between several building models in one session, with draggable inspection panels and WebWorker processing for large files. Ships with six sample IFC models across pharmaceutical, hospital, industrial, office, residential and clinic building types. A construction-sequence timeline feature exists in the codebase but its own roadmap lists it as not yet completed. Built between May and June 2025; dormant since, with no development in over a year.
A standalone, no-code pipeline orchestration engine. Workflows are built visually on a drag-and-drop canvas rather than hand-coded, backed by a versioned business-rule engine, role-based access control and multi-tenant organisation support. Wormwood runs as its own independent service, with its own API, database and authentication, and integrates with other systems - including the deltaPrism graph database - purely over HTTP, never as an embedded component of another product.
An optimized AI/ML/HPC platform designed to reduce cost overheads and democratize advanced computing. Built on Dell PowerEdge servers with NVIDIA A40 GPUs running in an immersion cooling environment.
A machine learning benchmark project demonstrating optimized ML performance across different hardware configurations. Showcases efficient GPU utilization for medical image scanning with modest compute resources.
An internal research project focused on Computational Fluid Dynamics (CFD). Uses high-performance computing to analyze fluid behavior in scenarios like immersion cooling environments.
A Computational Fluid Dynamics implementation using GPUs to accelerate performance. Runs on Nutanix's hypervisor with NVIDIA A40 GPUs in an immersion cooling environment.
A public website and hosting infrastructure for Akademia Mistrzow, a Gdynia-based youth sports foundation - GDPR-compliant contact form with AWS SES email notifications, Google Analytics and Search Console integration, and a 100/100 Lighthouse score across every category.
A schema-driven form engine for React applications. Forms are generated at runtime from a JSON schema rather than hand-built per screen, so a new data-entry interface is a configuration change instead of a development task. Supports conditional field logic, custom validation, role-based field variants and per-tenant theming from the same schema.
and 2 more
A hybrid deployment of large language models across both private cloud immersion clusters and AWS. Features fine-tuned Meta Llama 3 models with consistent performance across environments.
Read tracking for HTML decks, briefs and proposals. Upload a document or point at one you already host, send a tracked link, and see who opened it, which sections they actually read, and where they stopped - section-level dwell time rather than a single "opened" flag. Open source under AGPL-3.0, with a hosted service on a free and paid tier.
and 1 more
An energy management platform designed for Nordic commercial and industrial sites, built around high-frequency telemetry rather than monthly meter reads. The design covers second-by-second data capture from distributed locations, day-ahead consumption forecasting, and ruggedised edge hardware for harsh environments such as commercial kitchens. Currently a completed platform design with a working cost-modelling calculator; the platform itself is not yet built.
and 1 more
A self-hosted language model inference platform running on private GPU infrastructure, giving internal products model access without sending data to external providers. Four GPU nodes serve different models behind a single OpenAI-compatible endpoint, with load balancing, per-developer session management and rate limiting handled by a dispatcher layer.
and 2 more
A live knowledge-base hit-tracking service for UV's ROTC governance framework. Tracks how many times a known issue has been hit, by whom, and disseminates the solution once one exists - intended to replace and augment per-project Markdown KNOWN_ISSUES.md files with one central, queryable service. Matching is pure keyword scoring; no embeddings, no vector search.
and 1 more
A white-label-capable, multi-tenant licensing engine. It assigns licences to software packages scoped per client and per project, and brokers role-based access control decisions on behalf of each licensed package. Consuming products integrate through a frozen, generated OpenAPI contract and a dependency-light .NET client SDK rather than hand-rolled HTTP calls.
and 1 more
A reverse-booking tender and auction platform for services and goods. Rather than browsing catalogues and comparing static offers, buyers publish a structured request and invited vendors bid against it, then the buyer compares the bids and awards the winner. Supports sealed and open auction formats across hospitality, catering, transport, activities, goods and custom service categories.
and 1 more
An internal engineering portal that consolidates technical documentation and operational tooling behind a single interface. The Forge pairs a searchable document library with a catalogue of infrastructure management tools, and exposes both through a token-authenticated REST API so other systems can query them programmatically rather than through the browser.
and 1 more
A structured operating framework that governs how UV plans, builds, tests and ships software. It defines explicit entry and exit criteria for every unit of work, requires evidence before any work is declared complete, and enforces its own rules automatically through tooling rather than relying on reviewer discipline.
and 2 more
A concierge platform for groups travelling to Gdansk that covers the whole stay rather than individual bookings. Organisers submit dates, budget and group size once; the platform assembles accommodation, dining, transport and activities into comparable offers coordinated with local suppliers. Currently at working-prototype stage, with an AI concierge demo running against live activity inventory.
and 1 more
An AI hybrid cloud platform integrating on-premises and cloud infrastructure for machine learning workflows, with resource allocation built around cost, not just capacity.
Next-generation property graph database platform engineered with Flask, PostgreSQL, and Redis, delivering high-performance analytics with GPU acceleration capabilities. Features dynamic schema evolution and enterprise authentication systems.
Yale Geophysical AI Competition Platform designed for advanced seismic data analysis and machine learning applications, integrating geophysical modeling with competitive research environments for academic institutions.
Stanford RNA 3D Folding Project utilizing computational methods for predicting RNA secondary and tertiary structures, integrating machine learning with structural biology for molecular design and drug discovery.
A deployment of Unreal Engine 5 on an immersion-cooled server cluster. Optimized for rendering and visualization tasks with GPU-accelerated performance.