Forbes contributors publish independent expert analyses and insights.
Creative work starts messy. Briefs scatter across docs and boards. Reference links pile up. Teams bounce between prompts, files, and comments while a client expects a tight concept by tomorrow morning. New startup First Concepts wants to turn that chaos into a single place where AI facilitates the creative process, but “taste” stays firmly in human control.
Coming off a win at an AWS Startup Pitch Competition at the Upscale Conference in Málaga, Spain, the team showcased an approach that lowers the cost of building and testing creative systems for real teams. The idea is to enable these creative teams to move faster, accelerated with AI, without losing their creative style.
First Concepts cofounder Conor Hoey laid out the problem in plain language. “Creatives today are under intense pressure to deliver near final concepts in hours, not weeks.” . Pressure alone does not sink a project. Loss of context does. Briefs drift in Google Docs. Research hides in Perplexity. Inspiration sits in Pinterest or Cosmos. Outputs get assembled in Figma. That shuffle burns time and memory. Hoey answers those challenges with the startup’s approach that provides, “an AI-native workspace to help you go from brief to pitch-ready concepts up to 70% faster, whilst keeping your creative quality and taste intact. We call it ‘delivering taste at speed'” .
The core of the idea relies on fit with existing habits rather than a tool swap. Interacting with users through a browser extension versus another siloed app, “The system itself sits on top of your tool stack. It understands what you’re working on, unifying your team and briefs, as well as pulling relevant inspiration and research seamlessly. It then takes that context and orchestrates it around Gen AI tools so that you can focus on getting the best,” said Hoey at the pitch .
A layer that rides above the stack can carry brand DNA from one exploration to the next. That pattern reduces rework. It gives a team a memory.
The creative economy has an angst-filled relationship with AI. On the one hand, creatives see AI as a potential threat to their ability to earn a good paying living as well as a thief of their outputs, while at the same time seeing AI as a potential tool to help them accelerate and enhance their own outputs.
With AI the topic of every conversation, every tool in the creative chain now claims a dose of AI. From industry giants such as Adobe, Canva, Figma to smaller upstarts, AI, especially generative AI, is part of every creative process. The easiest path to AI adoption is to embed those capabilities within the tools and their existing user flows. Editing tools add generators. Design tools embed suggestion panels that sit next to layers and frames. Document tools learn to answer questions and draft outlines that draw on workspace memory. That strategy makes sense for large platforms with huge user bases. Keep people in the place they already know. Add one more power slider.
First Concepts pushes a different angle. Rather than providing a new surface to replace Figma or a clone of a board app, the company aims to provide an orchestration that sits above work in progress. The layer reads briefs, references, and prior choices. Then it routes a prompt to the right model with guardrails that match the brand.
The focus on process versus output for creatives is a compelling argument as it puts the human back in the center of the creative loop. To ensure its ongoing value, the company and its users will need to determine if the system truly learn taste, rather than just prompts. Brand DNA is not a style token. It lives in choices a team made over years.
Rejected approaches matter as much as final picks. A useful memory must capture that negative space. If the workspace only stores parameters or a library of prompt templates, it will drift back to a generic flavor, yielding a generic or average output that will quickly exude an AI aroma. Keeping the human taste in the center of the process is key.
Making sure to address where governance and rights land will also be critical. Large agencies ask hard questions about asset rights, model provenance, and audit trails. If an output is pulled from a dataset a client rejects, someone will need a record that shows exactly what happened. An orchestrator can help if it logs decisions and sources. It can harm if it turns the path into a black box. Buyers will ask for receipts.
The future of the creative industry will depend on how the industry adopts AI. The team’s makeup lines up with the problem space. The founders include a big-tech operator who also makes art. A brand and systems designer who understands the pain of scattered context, and a full-stack AI engineer. Advisors include large agencies and brand groups. Hoey framed it as a group with both craft and build skills.
He wrapped up his pitch with an eye to the future, “We’re building the infrastructure for creative work for the next decade. A living system that understands your creative DNA, predicts your workflow, and powers you to create the best.”
