How AI Accelerates Design at DockYard

Graphic designer workplace with computer pc, stationery and color swatches on wooden table.
Jon Akland

Associate Director of Product Design

Jon Akland

Building AI-powered applications requires the right expertise. From data engineering to model deployment, our team provides the end-to-end support you need. Book a free consult today to learn more.

Keeping up with the rapid proliferation of AI-powered design tools in the past 18 months or so has been a challenge, and I’ve been as skeptical as anyone of the bold claims they make. Some of the emerging features and workflows seem to be of dubious value, and many are struggling to differentiate from their competition; most are unproven to remain relevant during a remarkable moment of upset and norm-breaking in what was already a tumultuous period for the software industry.

At DockYard, our design and development teams have been applying a new set of tools to how we craft custom software, and it’s transformed the way we work and communicate. The biggest impact on our design process is apparent in two places so far: business development and early-stage design outputs.

Pre-Project Work

The discovery work we engage in to win client projects — arguably one of the most exciting stages in a project, when we’re meeting interesting people and the work is fueled by our natural curiosity to start something new — is unpaid and often complicated. The elements of every project are familiar, but no two are the same, and clients look to us to help them sort and prioritize all the intricate details. We’ve found that current AI systems can be very good at organizing and synthesizing project requirements and goals, as well as estimating team composition and timelines.

Time is always a factor while we’re assembling proposals for clients, but it’s important to get our facts straight before we estimate what resources we’ll need to complete a project successfully. Potential clients deserve to understand clearly the work ahead of us and what it will cost. We use custom GPTs — a structured set of prompts specifically written for this purpose — to define and organize project goals, constraints, and deadlines into a product requirements document (or PRD) that helps guide conversations with potential clients about scope, priorities, and approach. Clarity is crucial, and AI accelerates our understanding at this stage and facilitates accurate communication.

Early-Project Design Work

Design work that is systematic in nature or is primarily an artifact for project alignment and client/team discussion is ripe for AI assistance. In early project phases, wireframes and a variety of diagrams (userflows, sitemaps, Gantt charts, and swimlanes all have their place) serve to capture and illustrate product vision and feature priorities. The better we’re able to display and discuss these concepts, the more we can debate their merits and refine our intent.

We use another custom GPT to formulate a design specification document to structure our prompt before we ask AI-powered tools to render the visual format we need. After refinement, we can quickly get the work in front of engineers and clients to achieve alignment and clear the path ahead for engineering and design.

If diagrams get us to a set of shared project planning expectations quickly, the next place for AI to play a role is prototyping. If you’ve been in the industry for a few years, you’re probably used to the prototyping stage — whether design-only click-throughs or lightly coded proofs of concept — happening later in development, once product decisions have been established and engineering’s work is well underway. By dramatically lowering the bar to being able to achieve something real and functional without needing a great deal of human-authored design or code, prompt-to-prototype AI tools allow teams to jump ahead to prototyping as a means to experiment and test out feature priorities and product directions early in the process. Exploring multiple possibilities up front with minimal financial investment is an opportunity to de-risk the project for clients and DockYard’s project team alike. It’s also a great opportunity to engage clients in a collaborative exercise that has real value for their product, strengthening our business relationship.

Value of Design

Integrating AI tools into the design process at DockYard has proven so far to significantly accelerate and enhance our ability to understand and estimate potential projects during initial conversations with potential clients, whether they’re startups, growth-stage companies ready to scale, or established enterprise organizations. Once an engagement is underway, AI-powered design tools help us articulate project goals and strategic directions, and clearly communicate with teams and stakeholders — and speeding up and increasing the fidelity of early-stage artifacts allows our conversations around product decisions to be themselves higher-fidelity and more productive.

Once these initial conversations are held and the important decisions are made (at least for now), our creative design and development process can move swiftly and confidently thanks to a collective sense of clarity and purpose. That’s where the real fun begins.

Newsletter

Stay in the Know

Get the latest news and insights on Elixir, Phoenix, machine learning, product strategy, and more—delivered straight to your inbox.

Narwin holding a press release sheet while opening the DockYard brand kit box