Contact for workbachmatoan2910@gmail.comLinkedInAI automationPlatform engineering
mbach.dev
backend platforms, AI systems, and infrastructure-aware engineeringB.Eng 2026

I build automation that helps teams grow.

I am a Software Engineering student at Seneca Polytechnic focused on networking, servers, and backend platform development across cloud and on-prem environments. RHCE, RHCSA, CKAD, and Terraform Associate certified; hands-on with Linux, RHEL, Kubernetes, AWS, Proxmox, Docker, Jenkins, and GitHub Actions. I like building reliable services, automating infrastructure, and shipping AI-assisted systems with strong observability.

View ProjectsDownload ResumeGitHub
Toronto, ONResearch Assistant at Seneca Applied Research
Linux + AI + PlatformToronto
Systems I like building
Reliable services, developer tooling, and infra-aware products.
Portrait of Ma Toan Bach
Current focus
AI agent systems for testing and automation
Infra stack
Linux, Docker, Kubernetes, Terraform, CI/CD
Credentials
CKADTerraform AssociateRHCSARHCE
Current role
Research Assistant
Recent role
AI Engineer Intern
Certifications
4
Resume projects
8
Experience

Recent work in AI automation, firmware test systems, and internal platforms.

Most of my recent work has been around AI-assisted workflows: browser automation, firmware test generation, observability, and internal tools that help engineers and operators move faster.

Research Assistant

Seneca Applied Research
May 2025 - Present
Toronto, ON

Building AI-assisted firmware test systems with internal tooling, observability, concurrency controls, and close collaboration with platform teams.

  • Designed and deployed an automated firmware unit-test generation workflow using Python, LangChain/LangGraph, and FastMCP, cutting manual setup time by 90%.
  • Architected and containerized a microservices test-generation FastMCP server to expose 12+ test tools via an internal API platform, enabling parallel and sandboxed test runs with robust concurrency controls.
  • Implemented a RAG pipeline over 4,000+ code symbols and knowledge-base records, boosting coverage from 84% to 99% while reducing token usage by about 32%.
  • Integrated end-to-end observability with OpenTelemetry, Grafana, Tempo, and Loki, decreasing debugging time by 30% and improving reproducibility.
  • Partnered with AMD firmware engineering teams to refine prompts, guardrails, and platform features, improving AI-generated test quality and driving adoption across the organization.
PythonFastMCPLangChainLangGraphOpenTelemetryGrafanaTempoLoki

AI Engineer Intern

Pathway Communications
December 2025 - April 2026
Toronto, ON

Built AI automation tooling for browser workflows, internal observability, and agent-accessible data platforms in production-adjacent environments.

  • Built an AI browser automation system to extract and capture information from third-party portals using Playwright, TypeScript, OpenCode SDK with sandboxed execution, Jenkins CI/CD, and Docker for containerization, reducing manual coding time by 96%.
  • Assisted in testing AI automation workflows for ticket processing in the production OTRS system using the Google Agent Development Toolkit and Python.
  • Built a Django-based observability dashboard to support debugging and testing of AI automation workflows and collaborated with ITSM, DevOps teams, and stakeholders to gather feedback and align with project requirements.
  • Architected and maintained a FastMCP server platform that exposed database tools across MySQL, PostgreSQL, and SQLite for AI agents to query client information.
  • Integrated and configured Langfuse, Grafana, Prometheus, and OpenTelemetry to monitor LLM agent workflows and MCP server metrics, tracking cost, latency, token usage, and error rates to improve reliability and speed up troubleshooting.
TypeScriptPlaywrightDjangoFastMCPDockerJenkinsGoogle ADKLangfuseGrafanaOpenTelemetry
Certifications

Proof of hands-on Linux, Kubernetes, and infrastructure depth.

The quick cert chips are useful, but the real signal is what each certification covers. This section shows the operating systems, container, Kubernetes, and infrastructure-as-code skills each badge represents.

The Linux Foundation2025-01

CKAD

Demonstrates practical Kubernetes application skills: containers, orchestration, ConfigMaps, Secrets, and deploying cloud-native workloads.

Skills covered on the badge
Cloud-Native ApplicationsConfigMapsContainersDockerKubernetesOrchestrationSecrets
Verify credential
HashiCorp2025-10

Terraform Associate

Shows competence with infrastructure as code using Terraform: authoring configurations, using modules, managing state, and automating cloud infrastructure workflows.

Skills covered on the badge
Cloud ComputingInfrastructure as code (IaC)Infrastructure AutomationMulti-CloudTerraformTerraform CLITerraform ConfigurationTerraform ModulesTerraform State
Verify credential
Red Hat2025-09

RHCSA

Proves core Linux system administration skills in RHEL: filesystem and storage work, software/package management, networking, users/groups, Bash, security, firewalls, and SELinux.

Skills covered on the badge
Red Hat Enterprise Linux (RHEL)ContainerizationFile ManagementManage softwareBashLocal Storage ConfigurationFile System ConfigurationNetworkingSecurity ManagementBasic container managementSecurityFirewallsSELinuxUsers and Groups
Verify credential
Red Hat2026-04

RHCE

Validates hands-on Red Hat and Linux administration beyond the basics: system deployment, storage, automation, container workflows, security, and operational maintenance.

Skills covered on the badge
Software packages and repositoriesTask schedulingUsers and GroupsPodmanConfigure local storageCreate and configure file systemsDeploy, configure, and maintain systemsShell ScriptsManage security
Verify credential
Featured builds

Selected projects across AI products, backend systems, and infrastructure labs.

These are the builds that best represent how I work: shipping usable software while still caring about architecture, deployment, observability, and operational detail.

Overview

A solo interview-practice platform that combines a coding workspace with an AI voice interviewer.

  • Built a solo LeetCode-style interview practice platform with an in-browser problem workspace, Python editor, execution output, timers, and settings.
  • Integrated an AI voice interviewer using OpenAI Realtime to guide users through a structured interview flow covering approach, implementation, testing, and complexity review.
  • Implemented authentication with Auth0 and connected client-side code execution via Judge0, deploying the web app to Vercel with GitHub-based deployments.
Live DemoGitHub
/leetyap
01
Project snapshot
leetyap.com
TypeScriptReactNext.jsRedux ToolkitTailwindAuth0CodeMirrorOpenAI Realtime
Overview

A hackathon prototype for food-industry analytics, inventory visibility, and operational dashboards.

  • Collaborated in a 5-person hackathon team to prototype a food-industry analytics plus POS dashboard for inventory, sales trends, and near-expiry alert workflows.
  • Led the MVP front end in Next.js and React, shipping company vs. supplier views, KPI dashboards, certification workflow UI, and PDF report generation.
  • Implemented CI/CD with GitHub Actions to build and push Docker images to AWS ECR and deploy the container to Amazon EKS with Kubernetes manifests and ingress.
Live DemoGitHub
/datatrace
02
Project snapshot
datatrace
TypeScriptReactNext.jsDockerKubernetesGitHub ActionsAWS ECRAWS EKS
Overview

A Proxmox-based homelab for repeatable build-break-fix practice across Linux, networking, containers, and cluster operations.

  • Built and maintained a personal Proxmox-based homelab with 64GB RAM and dual Xeon E5-2660 v3 CPUs to practice infrastructure, systems, and automation through repeatable build-break-fix cycles.
  • Provisioned and rebuilt Linux VMs to practice Red Hat-style administration and troubleshooting across users and permissions, systemd services, storage, filesystems, SSH, and DNS and network fundamentals.
  • Planned and documented a compact OpenShift-on-Proxmox deployment using the agent-based installer, including static IP and DNS mapping plus Terraform-driven provisioning workflows.
  • Explored Kubernetes cluster operations and CI/CD experiments with Jenkins, capturing configs and lessons learned as a reusable knowledge base.
GitHub
/homelab-cluster
03
Project snapshot
Homelab
ProxmoxRHELNetworkingTerraformAnsibleKubernetesOpenShiftJenkins
Overview

A Go backend project that simulates core banking workflows with authentication, transactions, and CI-backed database testing.

  • Built a Go and Gin REST API simulating core banking workflows including user registration and login, account management, and money transfers, backed by PostgreSQL.
  • Implemented secure authentication with bcrypt password hashing, JWT middleware, session records, and account-ownership checks for protected resources.
  • Designed transactional transfer logic that records transfers and accounting entries and updates balances atomically, including stable ordering to reduce deadlock risk under concurrent transfers.
  • Added database-backed automated tests and GitHub Actions CI that starts PostgreSQL, runs migrations, and executes the test suite on every push.
GitHub
/very-simple-bank
04
Project snapshot
simple-bank
GoGinPostgreSQLsqlcJWTDockerGitHub Actions
Overview

A LinkedIn job monitoring bot that combines browser automation, AI filtering, storage, and Discord distribution.

  • Built a LinkedIn job monitoring bot that scrapes new internship and co-op postings and delivers shortlisted roles to Discord channels for student job hunting.
  • Automated LinkedIn login and UI-based job filtering with Node.js and Playwright, extracting job title, company, location, and apply URLs from dynamic pages.
  • Applied OpenAI-based validation to filter for CS-related internship and co-op roles in English before publishing to Discord.
  • Implemented deduplication with DynamoDB using a 14-day TTL to suppress repeats and broadcasted approved jobs to two Discord channels serving 144 students.
GitHub
/job-scraper
05
Project snapshot
LinkedIn Job Bot
JavaScriptNode.jsPlaywrightOpenAI APIDynamoDBExpressDiscord Webhooks
More from the resume

More systems work from the resume: compilers, protocol servers, and lower-level networking.

These projects round out the lower-level side of my work and show the same pattern as the main portfolio pieces: building from first principles and learning by shipping.

GoCompilerVirtual MachineGitHub Actions

turtle programming language (Go)

A small programming language in Go with both an interpreter-style evaluator and a bytecode VM.

  • Implemented a small programming language in Go, including a lexer, Pratt parser, AST, runtime object system, and built-in functions.
  • Built two execution engines: a tree-walking evaluator for interpreting the AST and a bytecode compiler that lowers AST nodes into instructions for a stack-based virtual machine.
  • Implemented VM features including globals, call frames, closures with free-variable capture, and symbol resolution across global, local, and builtin scopes, with unit tests across parser, compiler, and VM behavior.
GitHub
C++23POSIX Socketspthreadszlib

HTTP/1.1 Server

A lightweight HTTP/1.1 server in C++ built to practice sockets, request parsing, routing, and concurrency.

  • Built a lightweight HTTP/1.1 server from scratch in C++23 using TCP sockets, implementing request parsing, response construction, and routing for GET and POST endpoints.
  • Implemented basic file upload and download support using a configurable storage directory and appropriate success and error responses.
  • Added response compression and multi-client request handling to improve performance and scalability under concurrent traffic.
GitHub
C++23UDP SocketsDNSCMake

DNS Server

A lightweight DNS server in C++ for learning UDP protocol handling, binary parsing, and message serialization.

  • Built a lightweight DNS server in C++ that receives and responds to queries over UDP, implementing binary message parsing and serialization for DNS headers, questions, and answers.
  • Implemented domain-name encoding and support for compressed names using pointer-based labels to correctly parse real DNS packets and construct valid responses.
  • Enabled the server to forward queries to a public DNS provider and return the resolved results, validating end-to-end behavior using dig.
GitHub
Technical focus

The skills that keep showing up across the resume.

Across research, internship work, and personal projects, the pattern is consistent: Linux-heavy infrastructure, AI automation, backend services, observability, and tools that other people can rely on.

AI + Agent Systems

I build LLM-powered workflows, browser automation, internal tool servers, and observability around agent behavior.

FastMCPLangChain/LangGraphOpenAI modelsGoogle ADK

Platform + DevOps

I enjoy the infrastructure layer: Linux, containers, CI/CD, Kubernetes, AWS, and practical troubleshooting.

LinuxDockerKubernetesTerraformAWSGitHub Actions

Full-Stack Product Engineering

I also ship product-facing software with strong developer ergonomics, API design, and thoughtful UI execution.

Next.jsTypeScriptGoFastAPIREST APIgRPC
Education

Seneca Polytechnic

Bachelor of Engineering in Software Engineering

Sept 2022 - August 2026
Toronto, ON

Languages

PythonGoTypeScriptSQL

DevOps

LinuxRHELProxmoxBashDockerPodmanKubernetesAnsibleTerraformGitHub ActionsJenkins

Observability

LangsmithLangfuseGrafanaPrometheus

Cloud

AWS EC2AWS EKSAWS ECRAWS IAMAWS S3AWS RDS

Full-stack

FastAPIFastMCPOpenAPIREST APIgRPCGinReactTailwind CSSHTML

Databases / AI

LangChainLangGraphGoogle ADKChromaDBPostgreSQLMongoDBDynamoDBOpenAI models
Contact

Building something useful? Let's talk.

I am interested in new grad engineering roles across AI systems, backend and platform engineering, developer tooling, and infrastructure automation.

bachmatoan2910@gmail.comLinkedInGitHub