Sisyphe · Services Index Build Advise Teach Clients Contact

Build, advise, teach

Sisyphe is small on purpose. One senior engineer, with trusted collaborators when the work needs them, on problems where the model has to meet a real instrument, a real stage, a real signal or a real collection. If you are not sure your problem fits, write anyway.

01

Build

Working software in your repo, measured against the real environment it has to run in.

Agents inside creative tools

You get an agent that can operate your whole application, and an evaluation that tells you when its coverage slips.

Lanes of clips in a session, MIDI notes, audio waveforms and automation, with a bundle of fine violet lines threading through them from left to right. Stretches of a few clips are coloured blue where the path reads them and red where it edits them; a blue dot marks where the path is now. Lanes of clips in a session, MIDI notes, audio waveforms and automation, with a bundle of fine violet lines threading through them from left to right. Stretches of a few clips are coloured blue where the path reads them and red where it edits them; a blue dot marks where the path is now.
01aAn agent’s path through a session’s tracks and clips, reading some and editing others · drawn with Kiln

MCP servers and tool design for DAWs, plug-ins, creative apps and collections. We decide what the model should see and build the bridge into your object model or archive.

Our work
Sideman

Music, audio and voice ML

You get music and audio models that hold up in a product, not only on a benchmark.

A voice waveform on the left whose component waves fan out into separate frequency rows, become a hatched spectrogram in the middle, and end on the right as a grid of square tokens, each hatched according to its energy. A voice waveform on the left whose component waves fan out into separate frequency rows, become a hatched spectrogram in the middle, and end on the right as a grid of square tokens, each hatched according to its energy.
01bA voice waveform split into its partials, then a spectrogram, then discrete tokens · drawn with Kiln

Music source separation, audio classification and recommendation, speech recognition and synthesis. Data pipelines, training, evaluation, and getting a model into a product.

Our work
Roadie · Nimrod

Interactive art and installations

You get a real-time system that answers people, sensors or signals, with a mapping you can defend.

A room seen from above, filled with thousands of fine flowing strokes. Three small sensors with rings around them read one visitor, a blue dot; the strokes swirl around the sensors and part around the visitor. A room seen from above, filled with thousands of fine flowing strokes. Three small sensors with rings around them read one visitor, a blue dot; the strokes swirl around the sensors and part around the visitor.
01cThree sensors read one visitor and bend a field of strokes across a room · drawn with Kiln

Embodied AI for the stage and the gallery: real-time composition, sound, light and machines that respond to sensors, biosignals, movement or game state. Installations, performances, museum and gallery prototypes.

Our work
Dream Fauna (personal)
02

Advise

Read the paper, run the experiment, and say plainly what will work before you build it.

Research to product, fractional CTO, scientific advice

You get a senior second opinion, or your prototype turned into software, before you commit a team to it.

A lattice of triangles drawn as loose, repeated pencil strokes on the left that gradually straighten into clean double lines with ringed joints and a dimension line on the right. Two triangles are hatched in colour. A lattice of triangles drawn as loose, repeated pencil strokes on the left that gradually straighten into clean double lines with ringed joints and a dimension line on the right. Two triangles are hatched in colour.
02aOne structure, sketched loose on the left and built clean on the right · drawn with Kiln

For creative-tech teams and cultural institutions with a paper, a lab prototype or a collection that needs to become software. We talk to the scientists and ship the code, or sit beside your team as a fractional CTO or scientific advisor.

03

Teach

Workshops and courses for studios, schools and art programs.

Teaching and workshops

You get a team that can build and run agents without us.

Hundreds of short marks scattered at random angles on the left, each leaving a faint trail as it moves right and settles into place in a single spiral pattern, like the seeds of a sunflower. A few marks are coloured so their paths can be followed. Hundreds of short marks scattered at random angles on the left, each leaving a faint trail as it moves right and settles into place in a single spiral pattern, like the seeds of a sunflower. A few marks are coloured so their paths can be followed.
03aScattered marks, each an agent, finding their place in one shared pattern · drawn with Kiln

Sideman workshops for studios, schools and labs, from a day to a term, from “what’s in my Set?” to a finished track. Custom courses on agents for creative tools, audio machine learning and sensor-to-actuator systems.

Client work

Engagements for companies building AI into creative, cultural and voice products.

  • Tamada × Visto

    Fractional CTO · 2023–2025

    An AI curation engine for Visto, the art consultancy that builds collections for luxury hotels and landmark destinations: search by text or by image, taken from prototype to product with a team.

  • Persona

    ML engineer, embedded · 2023–2025

    Real-time conversational voice pipelines, built inside the product team.

  • Parts Order

    Scientific advisor · 2023–2025

    Scientific advice on music source separation.

For the two decades before Sisyphe, see the background.

How it works

  1. A written brief

    Half a page: what you are building, where it is stuck, what a demo would need to show. We reply with a scope, a number of weeks and a price.

  2. Two to eight weeks

    A fixed engagement with a clear question. Working code in your repo from the first week, a check-in each week, and no open-ended retainers.

  3. A demo, then a decision

    It ends with the work running in front of you and a written account of what worked and what did not. You decide what happens next.

Advisory
The other shape: a few hours a month to review architecture, models or an agent design before you commit a team to it.
Open source first
If the work belongs in Sideman or a public library, we will say so and price it accordingly.

Start a conversation.

Write with what you are building and where it is stuck. You will get a reply from the person who would do the work.

Write to us

Or copy the address: sisyphe.lab@gmail.com

Work with us