Applied AI Engineer · Programming Language Designer · Creator of OriLang
I build applied AI systems, programming languages, developer tools, and software products. My current work centers on OriLang and its compiler, including tools that help AI models write and reason about code more reliably.
Eric Malamisura is an applied AI engineer, programming-language designer, and founder. He created OriLang, a statically typed, natively compiled programming language, and is building its compiler and AI-oriented language tooling.
His work on OriLang spans type inference, value semantics, automatic reference counting, capability-based effects, integrated testing, and LLVM code generation. The project reflects his focus on making ambitious systems ideas practical for working developers.
Eric is the founder of Elucidsoft (founded in 2010), an independent software company. He also builds applied AI features, a GPU-accelerated terminal emulator, the cloudlayer.io and Upstat software products, and the WarpKit and OriJS frameworks used in those products. His work moves between AI, low-level systems engineering, and product development.
This is OriLang. Testing is not a library — it is a language keyword, enforced by the compiler. Whole-program type inference, value semantics with automatic reference counting, and native binaries through LLVM.
// Every function ships with its test — enforced by the compiler.
use std.testing { assert_eq }
type Shape = Circle(radius: float) | Square(side: float);
@area (s: Shape) -> float = match s {
Circle(radius:) -> radius * radius * 3.14,
Square(side:) -> side * side,
}
@test_area tests @area () -> void = {
assert_eq(actual: area(s: Circle(radius: 2.0)), expected: 12.56);
assert_eq(actual: area(s: Square(side: 3.0)), expected: 9.0)
}A programming language and compiler, a GPU terminal, two software products, and the frameworks that support them. Each case study explains my role, the technical approach, and the project’s current status.
Functional code. Imperative speed. Native binaries.
A programming language with mandatory testing baked into the compiler — statically-typed, natively compiled, no garbage collector and no borrow checker. Built for the age of AI.
A GPU-rendered, cross-platform terminal emulator.
A GPU-rendered cross-platform terminal emulator built from the ground up — every pixel drawn by a custom renderer at 60+ FPS, with a brutalist terminal-native interface.
HTML and URLs to pixel-perfect documents, at scale.
A dynamic PDF and image generation service — turn HTML, URLs, and templates into pixel-perfect PDF, PNG, and WEBP through a single API.
Operational intelligence for the teams who keep things up.
A collaborative incident-response and operational-intelligence platform for DevOps, SRE, and engineering teams — monitoring, on-call, status pages, and an AI assistant.
State-based routing for serious Svelte 5 applications.
A standalone Svelte 5 SPA framework with state-based routing, a predictable navigation pipeline, config-driven data fetching, schema-driven forms, and real-time WebSockets.
A NestJS-inspired framework for Bun — without the decorators.
A Bun-native backend framework — fast, type-safe, with first-class testing support. Dependency injection, guards, and interceptors through TypeScript interfaces, no decorators required. Powers production systems.
I work across applied AI, programming languages, systems, and software products. If that overlaps with a problem you are working on, send me a note.
Get in touch →