Matheus
Teixeira

Brazilian, with over 5 years delivering scalable, high-performance solutions for logistics companies. I also build and run my own digital products, as a way to grow beyond code.

Background

  1. 2024.05 — Now

    Grupo Autoglass

    Mid-level Developer

    Across the Product Platform, WMS and DRP teams.

    • Designed and automated the SKU Kanban, previously 100% manually managed in Excel spreadsheets and requiring 5 people.
    • Removed WMS addressing from the synchronous flow, saving warehouse staff hundreds of hours.
    • Worked on the DRP feature that redistributes idle stock between units instead of raising a new purchase order, recovering more than 10 million in stranded inventory.

    ProductHub (current)

    The company's central product platform, replacing the legacy system as the source of truth for the catalog: any system or API that needs product context calls the hub. I work on the catalog migration and on continuous sync, with jobs that detect drift between both sides and publish the events to Kafka topics.

    Product development Kanban board

    Designed and delivered the digital board that tracks 25,000+ SKUs, replacing a spreadsheet five people maintained by hand. The main table is updated by a .NET job (Quartz) that holds the business rules, and a materialized view built from it feeds the Tableau dashboards. I optimized the queries with CTEs. Updated daily, no manual data entry.

    WMS: async slotting

    WMS runs the warehouse, from receiving through slotting across the DCs. The slotting algorithm scans every location, weighing product dimensions and turnover to pick the best one, and it ran synchronously: some items took up to 5 minutes with the operator standing there waiting, in an operation that slots thousands of products a day. Working with the business team, I found the step didn't need to be synchronous, since the item still waited on someone to physically put it away after digital slotting. Refactoring the algorithm was the right fix, but it didn't fit the urgency. I restructured the flow around a FIFO Kafka queue, keeping only the location lookup synchronous and notifying the user by email on completion. The wait was gone for the operator.

    DRP: slow-moving stock transfers

    DRP handles store replenishment, forecasting demand so each store gets what it needs before it runs out. The company was sitting on thousands of slow-moving items and still buying those same SKUs whenever a location needed them. I worked on the feature that flags that stock and, instead of raising a new purchase order, triggers a transfer from the location holding it to the one that needs it.

    .NET · C# · Angular · Oracle · PL/SQL · Kafka · DDD · xUnit · Azure DevOps

  2. 2021.06 — 2024.05

    Benner

    Apprentice → Junior → Mid-level

    Three years on a logistics product, reaching mid-level in 16 months.

    • Technical reference for the squad, reviewing the team's code and backing technical and business-rule decisions.
    • Lead developer on a payment partner integration — I even travelled to the client's city, at their invitation, to understand the operation first hand.
    • Last major delivery: led a fully customizable travel logistics checklist app from planning onward, integrated with the company's other systems and used to issue more than 5,000 checklists a month.
    Mid-Level Developer 2022.10 — 2024.05

    BL Mobile — logistics checklists

    Built the app that digitized transport checklists, integrated directly with the core system and replacing a paper process. The checklists were fully customizable, supporting different question and answer types set up by analysts and answered by drivers. I worked on the mobile app end to end and on the API and its layered architecture. In use across several of the client's branches, issuing 5,000+ checklists a month.

    Stack — API: .NET 7, EF Core, DDD, RESTful API, unit tests. Mobile: Flutter, Dart, dio, provider, shared_preferences, shipped to Google Play.

    Tech lead for the squad

    I was the squad's technical point of contact, both for engineering and for the logistics domain. I reviewed the team's code and backed the other developers on technical decisions and on questions about the product's business rules, keeping code standards and unit test coverage consistent across our work.

    Payment provider integration

    Lead developer on the integration of a new payment provider into the system. I traveled to the client's site at their invitation to see the operation's pain points firsthand and build against them more precisely.

    Architecture and DevOps squad

    Worked on the architecture and DevOps/CI-CD squad with Jenkins, improving the team's internal tooling: release and update processes, and pipeline automation.

    Junior Developer 2022.03 — 2022.10

    Worked on features and improvements in the logistics product, including the payment provider integration I went on to lead as a mid-level developer. Contributed to the product architecture squad, improving the internal development environment: release process automation and database scripts. Reviewed the team's merge requests before they went to production. Refactored critical paths for performance in routines running over large data volumes.

    Apprentice Developer 2021.06 — 2022.03

    First contributions to the logistics product: fixes and improvements, working with TDD, SOLID and unit test coverage.

    .NET 7 · C# · Flutter · Dart · EF Core · DDD · Jenkins · TDD · SOLID

Education

Active Projects

  1. 01 / Scheduling

    Aura

    A scheduling and management app for barbershops. Live on both stores, with subscription plans and paying users.

  2. 02 / Language

    Diffluent

    English review treated as code review, for people in tech who won't settle for just being "understood."

  3. 03 / Content

    Postlight

    Turns your app's screenshot into a ready-to-post Instagram graphic. Built to make creating social content for Aura faster and easier.

About

Matheus Teixeira

These are systems with a lot of business rules and a lot of integration surface. The hard part is rarely writing the feature — it is understanding the operation behind it, and changing things without breaking what already runs.

Outside work I build and run my own products. Keeping one live taught me plenty that never shows up on a Jira card, mostly about what breaks after the deploy.

I use AI daily while developing, building agents and skills to speed up my own workflow, and I'm working on integrating it into the applications I build.

Contact