Martin Tejeda
Fractional Builder
About
I'm a fractional builder. I embed with teams to lead research, design, and development from concept to launch, delivering the work of a small studio as one accountable partner. My focus is turning ambiguity into structure, building scalable systems, and translating customer insight into product decisions. Alongside client work, I'm building AI-first tools to modernize how teams synthesize and use research.
Experience
Fractional Builder
Mawrs
·
2023-PRESENT
Lead research, design, and development as an embedded partner.
Take products from concept to launch without the overhead of a traditional agency.
Past work includes products for MDSV Capital, True-See, and SouthEast Bank.
AI Design process
Research Synthesis
After a round of user interviews, the team usually has to wait a few days for someone to go through recordings, pull quotes and write up findings. Rather than waiting, I upload interview files into Peridot and get structured insights, tagged quotes and highlight reels back in seconds. Instead of scheduling a readout meeting and hoping people read the doc, I send stakeholders shareable highlight reels with the raw data behind them. Product managers can explore the insights themselves. Engineers can hear users describe the problem in their own words.
Design Exploration
A cardinal sin teams commit after getting research results back is committing to one direction too quickly. I use Figma Make to leverage our existing designs and the latest research results to generate dozens of variations in minutes. Then I stress-test each one against the research insights the team already aligned on. When I bring options to a design review, I'm showing polished, opinionated design directions that are grounded in research and design principles. Reviews become much more effective.
Rapid Prototyping
Clickable prototpes in Figma don't cut it anymore. They provide surface-level feedback at best. I push finalized Figma designs directly into my codebase through Figma's MCP server, wire them up in Cursor and deploy a live prototype on Vercel. This process takes about 20 minutes on average, depending on the amount of screens that need to be brought into Cursor. As a result, users interact with real inputs, real flows and real data. The feedback is specific and actionable because participants aren't pretending to use the product. They're actually using it. Minor "papercut" issues that would have surfaced three weeks into development show up before a single engineering ticket gets written.
Development & Handoff
Clickable prototpes in Figma don't cut it anymore. They provide surface-level feedback at best. I push finalized Figma designs directly into my codebase through Figma's MCP server, wire them up in Cursor and deploy a live prototype on Vercel. This process takes about 20 minutes on average, depending on the amount of screens that need to be brought into Cursor. As a result, users interact with real inputs, real flows and real data. The feedback is specific and actionable because participants aren't pretending to use the product. They're actually using it. Minor "papercut" issues that would have surfaced three weeks into development show up before a single engineering ticket gets written.