Natalia Siwakowska joins FOTOhub as COO

Mateusz Ulewicz, Founder & CTO · · 6 min read

Natalia Siwakowska has been appointed Chief Operating Officer of FOTOhub, owning the 226-model catalogue and public API, workflows, and e-commerce integrations. Press release, platform goals, and a quote from her own announcement.

Natalia Siwakowska joins FOTOhub as COO
PRESS RELEASE

>

Bydgoszcz, 30 July 2026 — Natalia Siwakowska has been appointed Chief Operating Officer of FOTOhub, with responsibility for the development of the entire platform: product strategy, operations, the growth of the AI model catalogue and public API, partnerships, and expansion into e-commerce markets. The appointment is effective from July 2026.

>

FOTOhub currently offers 226 AI models through 267 documented REST endpoints. The platform is in the process of migrating fully to Amazon Web Services.

>

Media contact: [email protected]

Why an operations role, and why now

For its first years FOTOhub was an engineering problem. Providers had to be integrated, billing had to be built, and more than two hundred different models had to be made to behave like one product.

That phase is over. The platform now has 226 models, published SDKs in three languages, plugins in build for six store platforms, and customers processing catalogues numbering in the hundreds of items. That is no longer an engineering problem — it is an organisation, and an organisation has to be run.

Natalia Siwakowska works on precisely this moment. Her own description of her specialism:

"I step in when growth starts breaking how your team works."

It is hard to describe more accurately the challenge facing a platform that has been adding models and integrations faster than its processes could keep up.

Developing the whole platform

As Chief Operating Officer, Natalia is responsible for developing FOTOhub as a product and as an organisation — from strategy and priorities, through operations and process, to partnerships and entering new markets. In her LinkedIn announcement she set out four directions she is focusing on first.

More models, broader API

Currently 226 AI models — image, video, audio, speech, text and 3D — reachable through 267 REST endpoints across 35 modules. Alongside them, published SDKs: `fotohub` for Python and TypeScript, the `fotohubapp-cli` command line, and an MCP server with 30 tools, so an AI assistant can call the platform directly.

The goal for the coming months is simple to state and hard to execute: add models without breaking coherence. Two hundred models you cannot choose between is worse than ten that are well described.

Workflows that save time

Background removal, product photography, image editing. The point here is shortening the path from idea to finished file — so an operation that used to take half an hour in a graphics editor becomes one click, or one API call.

Some of those tools run on our own infrastructure, which is why they are free in the editor rather than metered per use.

E-commerce integrations

This is where the most work remains, and it is worth saying so plainly. The backend is live: the service handling bulk store jobs is deployed, supports nine job kinds and 63 ready-made presets across seven categories, with idempotency keys and signed callbacks.

Six plugins — Shopify, WooCommerce, PrestaShop, Magento 2, Shoper, BigCommerce — are written but not yet published to their marketplaces. Anyone running a store on one of those platforms who wants to be first can connect directly through the API today, without waiting for a listing review.

Powerful AI, made simple

The hardest of the four, because it cannot be measured in commits. Building access to two hundred models is an engineering task. Making sure the person running a ceramics shop knows which one to pick is a product and operations task.

Our answer is an orchestration layer: describe what you want to achieve, and the platform picks the studio, picks the model, and writes the parameters.

Partnerships and infrastructure

The catalogue is built with model providers ranging from large labs to specialised engines for narrow jobs. We deliberately do not try to replace all of them with our own models; some work calls for a tool we would never build.

We are in the process of migrating the platform fully to Amazon Web Services. It is one of the operational priorities for the coming months: one cloud instead of two, which simplifies both billing and the answer to a customer asking where their data lives.

We pay for AWS Business Support and take part in the AWS Activate programme for startups. On a platform where customers process entire product catalogues, support with a defined response time is not about prestige — it is about predictability.

From Natalia

From her announcement:

"Some dreams feel far away. Until one day you realise they have become your everyday life."

>

"I am lucky and proud to be part of a team building tools that make the best AI models easy to use for creators, businesses and e-commerce."

>

"And honestly? We are only getting started."

That last line is, at this moment, the most accurate sentence in the whole post.

Natalia Siwakowska — Chief Operating Officer, FOTOhub

[email protected] · LinkedIn · appointment announcement

Mateusz Ulewicz — CTO and founder of FOTOhub

LinkedIn

More about the platform: our models · API and console · integrations