BrainGrid
Builder Stories

Builder Story: Saymon and Core Care

From solo consultant to shipping faster than teams: How BrainGrid gave an 8-year engineer the PM he never had.

Nico Acosta
8 min read
Builder Story: Saymon and Core Care

"The feeling I have when using BrainGrid is that I have a PM with me."

  • Saymon
Video

#Background: The Senior Engineer Gone Solo

#Who is the builder?

Saymon is a software engineer from Brazil with eight years of professional experience. He has led teams of eight at top agencies, shipped features for apps with more than 9 million daily active users, and taken products from concept to seed funding. For the past two years, he has worked as an independent consultant, helping companies launch MVPs, secure funding, improve code quality, and develop growth strategies.

Saymon is not a vibe coder learning to build. He is a senior engineer who already knows how to build, and who recognized that the bottleneck was never the code itself.

#The Shift to Agentic Development

About a year and a half ago, Saymon made the transition to an agentic workflow. He describes it like the jump from JavaScript to TypeScript — a J-curve where the upfront investment pays off dramatically.

His conviction was simple: writing code is not really what a software engineer should be doing. Solving problems through software is. When tools like Cursor and Claude Code matured enough to handle the writing, Saymon shifted his energy to where it mattered most: research, architecture, and decision-making.

#His Development Setup

Saymon's current stack is built for speed and cost-efficiency:

  • IDE: Cursor
  • Primary coding agent: Claude Code (Opus 4.5 for most tasks, Sonnet 4.5 for large-context work requiring the million-token window)
  • Research: Gemini Deep Research as his go-to, supplemented by Claude Code research and GPT Deep Research
  • Planning and specs: BrainGrid

He still reads and reviews every line of code. But he no longer writes most of it.

#Vision: Taking Care of the People Who Take Care of Others

#What is he building?

Saymon is building Core Care, an all-in-one application for therapists in Brazil. The platform covers patient management, financials and payment handling, telemedicine, and one-on-one video sessions. As Saymon describes it, it is essentially seven startups consolidated into a single product.

#Who is it for?

Core Care targets therapists in the Brazilian market first, with a long-term vision to go global. Saymon describes the mission clearly: "Take care of the people who take care of others."

#What does it solve?

Therapists currently juggle fragmented tools for scheduling, billing, patient records, and telehealth. Core Care brings everything into one platform so they can focus on their patients instead of their software stack.

#Long-term vision

From day one, Core Care has been architected with internationalization and globalization built in. Saymon is not building a local tool. He is building a platform designed to scale globally from its foundation.

#The Turning Point: Choosing BrainGrid

#The pain point: Product thinking without a product team

As a solo consultant working across multiple client projects, Saymon faced a recurring problem. He had access to great product management tools, but they had no awareness of the codebase. He had powerful coding agents, but they did not understand product context. He was the bridge between the two, and that bridge was expensive.

When working across different projects, he would have to context-switch constantly. If he needed to understand how something worked in a project that was not on his machine, he had to ask another developer to research it and report back. The gap between planning and engineering was real, and it cost hours every week.

Saymon tried building his own system. He tried multiple existing solutions. None of them bridged both sides — product management context and codebase awareness — in a single tool.

Then he found BrainGrid, joined the waitlist, and started using it.

#What made BrainGrid different

BrainGrid gave Saymon something no other tool could: the ability to do requirements planning and task breakdown with knowledge of the codebase. It combined the product management layer with engineering awareness so he no longer had to be the translator between the two worlds.

"It is the same as having all the tools that I need for product management and also all the tools that I need for understanding at least the code base."

#How BrainGrid Changed His Workflow

#From organizing to decision-making

The biggest shift was not speed. It was where Saymon spent his time. Before BrainGrid, a large portion of his hours went to organizing: writing tasks, defining acceptance criteria, structuring plans. Necessary work, but as Saymon admits, "sometimes boring."

BrainGrid removed the boring part. What remained was pure decision-making. Saymon now spends his time evaluating options, refining architecture, and making calls. Not formatting Jira tickets.

"It removes the boring part. It's just decision-making. Okay, yeah, let's do things this way."

#His BrainGrid workflow in practice

Saymon's process follows a clear pattern:

  1. Understand the problem. Whether it is a client project or Core Care, he starts by deeply understanding the need and the problem being solved.

  2. Research. He studies how other companies and apps solve similar problems, cross-referencing with his own experience.

  3. Architecture. He sketches the high-level approach, including what to build and in what order.

  4. Spec with BrainGrid. He brings his architecture into BrainGrid and asks it to help create specifications. BrainGrid solves the blank-page problem and produces a first draft that Saymon then reviews and refines.

  5. Task breakdown. BrainGrid generates granular tasks with acceptance criteria. Saymon uses BrainGrid's MCP and CLI within Claude Code to pull tasks directly into his agent.

  6. Pre-task enrichment. Before executing, Saymon layers additional context onto each task. That means Supabase checks for database tasks, design system consistency prompts for UI tasks, and API validation for backend routes.

  7. Test flow generation. Based on the tasks, Saymon uses BrainGrid to generate the complete set of test flows he will need for QA. One recent feature produced 25 tasks and 13 distinct testing flows.

  8. Implementation. He runs each task through Claude Code, reviews the output, and moves to the next.

  9. QA and review. AI-assisted code review catches issues and bugs, followed by manual QA against the flows generated earlier, then automated tests.

#Outcomes: Faster, More Organized, Higher Quality

#A client feature that generated $1 million in revenue

One client came to Saymon with a hard deadline for a feature that would enable a major event. The challenge: eight interconnected apps that needed to work together, with architecture decisions that would determine long-term maintainability.

Without BrainGrid, the research and planning alone would have taken two full days of six to seven hours each. With BrainGrid, Saymon had a working plan in 30 minutes. He spent the rest of the day refining it.

The result: a feature that would have taken three to four weeks was delivered in just over two weeks. The client used it for an event that generated $1 million in revenue.

#A side-by-side quality comparison

Saymon ran his own experiment. He built the same SaaS MVP twice:

  • Without BrainGrid: Two days to a working MVP using Claude Code with direct prompting. The code was messy, lacked structure, and would require significant refactoring to maintain.

  • With BrainGrid: Four extra hours of upfront planning, then a similar two days of coding. But the result had more features, a cleaner architecture, and code that was far easier to maintain and extend.

The difference was not just speed. It was confidence.

"I move with confidence on my own code."

#The feeling of having a PM

For Saymon, the most meaningful change is not a metric. It is a feeling. As an individual contributor, he finally has the experience of working with a product manager — someone (or something) that handles the tedious but critical work of organizing requirements, writing acceptance criteria, and maintaining structure.

For his consulting work, the impact extends to teams. BrainGrid closes the gap between product managers and engineers, making both more productive. It is not replacing either role. It is making the handoff seamless.

"What BrainGrid is helping with is closing the gap between the engineers and the product managers and making both of them more productive."

#What He Wants Next from BrainGrid

Saymon's feature request is focused on adoption and accessibility:

IDE integration without requiring GitHub access. When working as a consultant across multiple client projects, getting GitHub access for each organization adds friction. Having BrainGrid integrated directly into the IDE — similar to how Cursor or Slack adopted bottom-up strategies — would make onboarding new engineers and new projects dramatically easier.

It is the kind of feedback that comes from someone managing multiple teams across multiple codebases, where every step of setup friction multiplies across every project.

#Final Thoughts

Saymon's story is not about learning to code with AI. It is about a senior engineer who already knew how to build and found the missing layer that let him operate like a full product team.

With BrainGrid, he is not just building faster. He is building more organized, more maintainable software — and shipping features that drive real business outcomes.

Core Care is just getting started, but the foundation is built for global scale. And Saymon has the workflow to match.

Check out Core Care, coming soon to therapists in Brazil and beyond.

Get Started

Ready to build without the back-and-forth?

Turn messy thoughts into engineering-grade prompts that coding agents can nail the first time.

Re-love coding with AI