Do I need to know AI to work with Dominika Kuś?

No. You tell Dominika what you want to achieve and she picks the tools. She explains along the way only as much as you need to make decisions. The rest is handled.

I only have a rough idea, not a finished brief. Can we still work together?

Yes. The discovery call exists partly to turn a rough idea into something concrete. You don't need to arrive with a finished brief - Dominika maps the pipeline based on your goals.

Can I keep using the AI system after delivery without additional fees?

Yes. The system is built for you once - then you use it however you want. No monthly licenses, no time limits, no SaaS. The system lives with you and works as long as the underlying tools (Claude Code, Higgsfield, Fal.AI etc.) keep running.

What sets Dominika apart from an AI agency or another freelancer?

Dominika builds systems, not one-offs. Plus a psychology background (MSc in Social and Organizational Psychology, SWPS University 2025) - every project starts with the question of how people make decisions, then tools are chosen. Solo operator, no subcontractors.

Can I start with something small to test the working relationship?

Yes, that is the recommended path. A single production (one film or one photo shoot) is a good way to see how Dominika works before committing to a bigger system. No long-term lock-in.

What is Dominika Kuś's stack?

Orchestration: Claude Code. Visual production: Higgsfield Super Computer, Fal.AI, Weavy. Video models: Kling 3.0, Sora, Runway, Veo 3.1, Seedance 2.0. Image: NanoBanana 2 Pro, Midjourney, FLUX, Seedream. Audio and avatars: ElevenLabs, Suno AI, HeyGen. Automation: Make, n8n.

AI Creative Systems Architect

I build AI systems that work for your brand 24/7

Multi-LLM systems in Claude Code. Higgsfield Super Computer and Fal.AI for image and video automation. End-to-end content automation, from brief to published asset.

Meet my robots.

Send a brief (one minute)

See the results

Robot Dominiki - geometryczna głowa z cyjanowymi oczami

Claude CodeHiggsfield Super ComputerWeavyKreaFal.AIElevenLabsKling 3.0SoraRunwayMidjourneyFLUXSeedreamNanoBanana 2 ProSuno AIHeyGenMaken8nVeo 3.1Seedance 2.0Pixverse30+ AI toolsClaude CodeHiggsfield Super ComputerWeavyKreaFal.AIElevenLabsKling 3.0SoraRunwayMidjourneyFLUXSeedreamNanoBanana 2 ProSuno AIHeyGenMaken8nVeo 3.1Seedance 2.0Pixverse30+ AI tools

01 / What I build

Three offers. One engine: Claude Code.

Pick the depth you need. Each path runs on the same stack, so you can start small and scale into a full system.

★ Flagship

AI systems in Claude Code

One engine. Three modes of use.

Multi-LLM pipelines (Claude + GPT + Gemini + 30+ specialised models)
API integrations with your stack: CRM, store, social, analytics
Visual production via Higgsfield, Fal.AI, Weavy, automated end-to-end

Tell me what you want to build

Single delivery

One production

Brief in, file out. No system.

One film, one photo session, one campaign. Delivered as a single project with 2 revision rounds.

Get a quote

Brand persona

AI avatar that doesn't sleep

Your face. Your voice. 24/7.

Digital brand personas with consistent look, voice and behaviour across every piece of content.

Build my avatar

02 / Showreel

See what I build

Each of these is a different pipeline. All orchestrated from Claude Code.

Film

▶

Zuzanna - AI Avatar

Zuzanna's avatar runs on a Claude Code system that remembers her character, personality, look and voice. It holds a stored library of references and prompts for every element.

Film

▶

Character and scene consistency

Claude Code can generate and remember the look of environments and characters, then produce a coherent film. Every shot lines up with the next.

Reel

▶

Complete AI scene

I generated each character, built a reference set, and synthesised their voices. Consistency held across scenes - look, voice, behaviour, and the environment itself.

🎁GIFT

A free workflow - click and try.

My self-prompting product photography workflow in Weavy. I build systems like this for clients. This one is public. Open it in your browser. No email. No install. No catch.

01
Drop in any photo of your product
Phone snap, bad lighting, whatever. The workflow handles it.
02
It generates your product from different angles
And upgrades the details to pro-studio quality.
03
It asks about your brand style and shot type
Lifestyle? Hero shot? Flat lay? Mood? A few questions, nothing complicated.
04
It writes intermediate prompts itself
Combines the info, decides on lighting, composition, mood, color grade.
05
Outputs a final prompt and final photo
File ready to publish. In minutes, not weeks.

One source photo → a full product session in your brand's style. No photographer. No shoot plan. No four-week wait.

▶Try the workflow (free, in browser)

PS: I still use it for every product session. It works.

04 / Process

How I work - 4 steps

Every project: same process. Call, system design, build, handover.

Discovery call

30 min, Google Meet or phone, free

We talk about your business. I ask about processes, KPIs, bottlenecks. Together we decide if the project makes sense.

System design

3-5 days

I map the pipeline. Pick the AI models. Design integrations. You get a fixed-price quote - not hourly.

Build

2 revision rounds included

I build in Claude Code, Higgsfield Super Computer, and the rest of the stack. I show progress.

Handover & support

post-launch

Documentation, training for your team, post-launch support. Optimization based on early results.

The discovery call is free. If after 30 minutes you see it's not for you - no problem. You only pay when we start the build.

05 / Paths

Two ways we can work

One if you need an asset now. The other if you're building for the long run.

A. Single delivery

You need an asset for a specific campaign, shoot, or launch.

Get a quote

B. A system built for you★

You're playing the long game. You need infrastructure, not one-off creatives.

Full pipeline in Claude Code (yours, not SaaS)
Integrations with your stack (CRM, store, social, analytics)
Documentation + team training
Post-launch support + optimization

Tell me what you want to build

The discovery call is free. Send a brief - you get a fixed-price quote. If I see we can't hit your goals, I'll tell you straight.

06 / Stack

Stack - 30+ tools in one pipeline

Claude Code orchestrates. Higgsfield Super Computer delivers image and video. Everything else executes specific jobs.

Center #1

Claude Code

Pipeline orchestration, automations, multi-LLM, RAG, integrations.

Center #2

Higgsfield Super Computer

Cloud visual production. Characters with continuity, dialogue, long scenes.

Video

Kling 3.0SoraRunwayMinimaxVeo 3.1Seedance 2.0LTXPixverseOmnihuman

Image

NanoBanana 2 ProMidjourneyFLUXSeedreamHiggsfield SoulGPT Image 2

Audio & avatars

ElevenLabsSuno AIHeyGen

Automation

WeavyKreaTapnowMaken8n

Design

CapCutFigmaFramerGamma

07 / About

About me

Dominika Kuś. MSc in Social and Organizational Psychology (SWPS University, 2025). AI Creative Systems Architect since 2024.

I started with psychology - to understand how people think, decide, and buy. Then came AI.

The two overlap in every system I build.

08 / FAQ

Questions I hear most often

No. You tell me what you want to achieve - I pick the tools. I explain along the way only as much as you need to make decisions. Leave the rest to me.

09 / Brief

Tell me what you want to build.

Short 5-step brief. You get a reply within 24h.

Step 1 of 5