Agentic SDLC consulting · Hands-on engineering
I help engineering teams put AI agents to work across the software development lifecycle. Starting with your codebase and how your team works, we choose a useful pilot, validate it, and plan a practical rollout.
Your team owns the decisions and the code.
02 / Agentic SDLC consulting
People, process, and codeFor teams that want a clear way to use coding agents in the software they already build and maintain.
The starting point is your delivery process: how work is defined, how your codebase is structured, and how changes are tested, reviewed, and released. Tool selection and agent setup follow from that context.
Review the codebase, architecture, tests, CI, documentation, and team handoffs. Find where agents could help and where engineering groundwork comes first.
A current-state assessment, prioritized opportunities, and a scoped pilot recommendation.
Build a bounded workflow on a real task, such as test coverage, an integration change, or a legacy refactor. Put repository context, instructions, checks, and human review in place.
A working pilot, reusable setup, and results compared with an agreed baseline.
Turn what worked into a repeatable practice. Define ownership, review expectations, team guidance, and the next use cases worth trying.
A practical playbook, team walkthrough, and rollout recommendations based on the pilot.
Agree on data access, allowed actions, test requirements, and the decisions that stay with a person. Keep production changes behind your team’s approval process.
Choose measures that fit the work: cycle time, review effort, rework, defects, or test coverage. Use the pilot to decide what to improve, expand, or stop.
Built for scale.
Grounded in delivery.
Daily orders processed by
SKULabs sync services
Historical orders in
Amazon migration workflows
Developer productivity from a
ClassWallet infrastructure fix
03 / Hands-on engineering
Implementation when the work calls for itConsulting grounded in building software. I also take on focused development work across integrations, SaaS, modernization, and AI-enabled features.
Make your systems work together.
Reliable data pipelines, partner integrations, and migration workflows that connect the platforms your business relies on.
API integrations · ETL · Data mappingTurn a business problem into working software.
Full-stack products and internal tools, from configurable workflows and billing infrastructure to custom CMS and CRM platforms.
React · Node.js · Business workflowsMove forward with the product you already have.
Untangle legacy systems, modernize application layers, and strengthen APIs, release practices, and developer infrastructure.
Architecture · Refactoring · PerformancePut AI to work on a specific problem.
Build AI-assisted data mapping and templating, prototype agentic workflows, and apply AI to development with clear boundaries and engineering review.
LLM features · Automation · Prototyping04 / Selected experience
Real systems. Real constraints.The production engineering behind my consulting approach, from technical leadership at ClassWallet to integrations and modernization at SKULabs.
High-volume order synchronization and large-scale seller migrations.
Keep commerce data moving as merchants grow, APIs evolve, and large amounts of historical data need to enter the platform.
Optimized Node.js, Redis, and MongoDB sync services using concurrency and incremental strategies. Migrated the Shopify integration to GraphQL with versioned rollouts, and built WIX and TikTok integrations from the ground up.
Sync services processed 100,000+ daily orders. Separate Amazon seller migration workflows handled 1M+ historical orders.
Configurable services, partner data exchange, and developer infrastructure.
Support new platform workflows, exchange partner data reliably, and remove a development bottleneck slowing the team.
Architected a configurable data-mapping utility, designed and led a survey service, and built a high-throughput ETL sync service for the FACTS integration.
The survey service reduced dependency on third-party tooling. Resolving a major Docker container bottleneck improved developer productivity by roughly 2×.
Architecture, incremental modernization, and AI-assisted development.
Turn legacy platform constraints into actionable work while keeping product and engineering aligned.
Authored API and domain modernization RFCs, established release and versioning practices, and helped lay automated testing foundations. Advanced React and Next.js modernization at SKULabs.
Led technical reviews and cross-functional planning, and prepared proofs of concept for AI-assisted modernization workflows with an emphasis on maintainability and confidence in changes.
These examples and figures reflect my professional experience as an employee. Pilot goals and measures are agreed for each engagement.
05 / The person behind the work
I’m Matheus, a full-stack engineer and technical lead based in Fort Lauderdale, Florida.
I enjoy the messy, high-impact part of software: finding the root cause, making trade-offs clear, and turning an ambiguous problem into a solution a team can use and maintain.
My background spans engineering leadership, technical product management, and hands-on development. At ClassWallet, my work includes technical leadership and a plan and proof of concept for agentic legacy modernization. I bring that implementation perspective to how teams adopt AI.
See my full background on LinkedInSoftware Technical Lead, AI-Driven Development
Previously Software Engineer II
Lead Software Engineer
Previously Full Stack Developer
Technical Product Manager & Full Stack Developer
Full Stack Developer & Web Developer
Also building
An independent builder’s perspectiveA subscription analytics platform bringing revenue, retention, churn, and payment insights into a clearer view for business teams.
Explore PozialA desktop interface I built for agentic project planning, development, shipping, and scaling. Released as a free macOS application.
Explore Mosaic06 / Working together
Clarity before code. Ownership through delivery.For an agentic development engagement or a software build, we agree on the outcome, scope, and review points before implementation.
Bring the problem, your current stack, and the delivery friction you want to address. We identify a useful starting point and the access needed to assess it.
Define the deliverables, success measures, responsibilities, and review checkpoints. Choose an assessment, a focused pilot, or a bounded development project.
Keep decisions and progress visible, validate the important workflows, and leave documentation and working examples your team can maintain.
The engagement covers how the team works with agents: task definition, codebase context, testing, review, and release practices. Tool and harness setup are part of the implementation when they support the workflow.
We choose a bounded task and agree on a baseline and success measures before building. The review considers the quality of the change and the effort required to verify and maintain it, alongside delivery time.
Yes. My experience includes legacy platform analysis, React and Next.js modernization, API redesign, and development-infrastructure improvements. We can start by reviewing the system and identifying a bounded, useful piece of work.
Engineering teams exploring AI-assisted delivery, teams already using coding agents but missing a consistent workflow, and teams modernizing an existing product are a good fit. I also build SaaS features, internal tools, integrations, and data pipelines.
Share your team’s goals, stack, current development process, and where work gets stuck. Include any AI tools you already use and your timeline or constraints. A high-level overview is enough for the first conversation.
Let’s talk about your team
Tell me how your team builds software today and what you want to improve. We can find a focused place to start.
Start a conversation on LinkedInA short overview of your team and workflow is enough.