Skip to content

Independent Research & Engineering

Engineering what comes next.

We investigate the foundations of intelligent and computational systems — turning research into technologies that expand what software can achieve.

A seeded vector field in which four streamlines, starting from separate corners, converge on one point. x y
Fig. 0 Four streamlines, seeded in separate regions of one field, converge on a shared point. Computed at build time; no data is represented.

01 Research philosophy

One question, investigated from four directions.

Our research is centred on a single question: how can we make complex computational systems more efficient, capable, and accessible? We are not doing everything. We investigate a focused set of connected problems and build systems to explore them.
01

Efficient

Computation is spent where it changes the outcome. We look for work that can be avoided, reused, or scheduled better, and we measure it honestly.

02

Capable

Systems that can act on underspecified goals, in complex environments, without silently guessing, and whose actions can be reviewed and reversed.

03

Accessible

Capability that is not limited to one hardware stack, one language, or one level of expertise.

02 Research domains

Four domains, connected by shared problems.

Each domain is a different angle on the same question. The connections between them are specific: every pair shares a concrete problem that neither can solve alone.
Rendered 3D scene: four dark ceramic blocks — Intelligence Systems, Computing Infrastructure, Simulation & Real-Time Systems, Applied Research Systems — stand on a glowing orbit around a luminous core in a glass sphere. Light filaments join every pair of blocks; each is a shared problem: Cost of computation (Intelligence Systems and Computing Infrastructure); Agents in tools (Intelligence Systems and Simulation & Real-Time Systems); Ambiguity and planning (Intelligence Systems and Applied Research Systems); GPU pipelines (Computing Infrastructure and Simulation & Real-Time Systems); Reproducible measurement (Computing Infrastructure and Applied Research Systems); Developer experience (Simulation & Real-Time Systems and Applied Research Systems).
Fig. 1 Four domains on one orbit around the core question. Every pair is joined by a filament — a problem both work on. Rendered in Blender from the same data as this page.

Core question

How can we make complex computational systems more efficient, capable, and accessible?

Every pair of domains shares at least one concrete problem. Select a domain — or a problem on the figure — to see what connects them and which projects work on it.

  • Cost of computation
  • Agents in tools
  • Ambiguity and planning
  • GPU pipelines
  • Reproducible measurement
  • Developer experience

04 Current investigations

What we are trying to find out.

Open questions, phrased so that they can be answered, with the stage of evidence each has reached and the next concrete step.
  1. Q01

    Is a meaningful share of language-model inference redundant, and can it be detected cheaply enough to skip?

    Evidence stage: Research hypothesis

    ZERO Research Lab · Next: Literature taxonomy and measurement protocol

  2. Q02

    Can an open model be taken from a model identifier to a validated run on ROCm reproducibly, with every measurement accounted for?

    Evidence stage: Prototype

    ROCmHub · Next: First run on physical AMD hardware

  3. Q03

    Can hardware ray-traced visibility be integrated into an existing forward renderer without regressing its baseline?

    Evidence stage: Prototype

    AETHRA Engine · Next: Integration into forward directional-light shading

  4. Q04

    How should an editor expose a live scene to an automated agent so that every change is previewed, reviewable, and reversible?

    Evidence stage: Prototype

    AETHRA Engine · Next: Wider operation coverage under the same transaction guarantees

05 Research methodology

Measured claims, stated limits.

We distinguish between a hypothesis, an experiment, a prototype, a validated result, and a released technology, and we label every project accordingly.
  1. 01

    State the hypothesis

    Write down the idea in a form that could be shown false, before building anything.

  2. 02

    Fix the measurement

    Choose metrics, baselines, hardware, and acceptable variance in advance, so results cannot be selected afterwards.

  3. 03

    Build the instrument

    Build the smallest working system that can test the idea under realistic conditions.

  4. 04

    Report what happened

    Publish the outcome with its limitations, including negative results and measurements that could not be taken.

Evidence stages

  1. 01 Research hypothesis A stated, falsifiable idea. No experimental evidence yet.
  2. 02 Active experiment Work is running to test the idea. Results are incomplete and unpublished.
  3. 03 Prototype Working software exists and is tested internally. Its central claims are not yet validated.
  4. 04 Validated result A result reproduced under a documented protocol, with the method and data available for review.
  5. 05 Released technology Validated work, publicly available and maintained for use by others.

Definitions and rules: How we report progress.

06 Open research

Notes now. Publications when there is something to publish.

Aporen has not yet published peer-reviewed work. We publish research notes as work progresses — methods, positions, and results, including negative ones.

07Collaboration

Working on a related problem?

We are interested in hearing from researchers and engineers working on inference efficiency, agent reliability, accelerator tooling, or real-time systems — especially if you can share hardware access, data, or a hard problem.