Hiring teams
Share role scope, team size, and expected start date by email or LinkedIn.
Helping teams build with confidence
Reliable systems that help your business move forward
I build secure, dependable systems and automate repetitive work, so your team can spend less time fixing problems and more time serving customers.
❯whoami
senad — devops & cloud engineer · aws certified
❯terraform apply -auto-approve
Apply complete! Resources: 128 added, 0 changed, 0 destroyed.
❯./deploy.sh --env production
✔ pipeline green · zero-downtime release
❯▊
Hiring teams
End-to-end platform ownership across architecture, delivery, and day-2 operations.
Freelance clients
Production-ready foundations with practical automation and clear, transparent delivery communication.
Shared outcome
Faster releases, fewer operational surprises, and maintainable cloud systems.
Projects
Cloud platforms, delivery automation, and open-source tooling, ordered by recency. Filter the complete collection by work type or stack.
DrawMe Natural-Language Diagramming for Pi
Built a Pi extension that turns natural-language descriptions into validated, editable draw.io diagrams and local image exports.
- Pi
- TypeScript
- Node.js
- npm
- +2 more
Details
- Problem
- Architecture and workflow diagrams are valuable documentation, but manually laying out editable diagrams and checking connections, geometry, and export quality interrupts development flow.
- Approach
- Built DrawMe as a Pi extension that plans diagrams from natural-language prompts, authors draw.io XML or Mermaid, validates each revision with a deterministic structural linter, reviews rendered previews when the active model supports images, and exports editable PNG, SVG, PDF, or JPG deliverables through the local draw.io CLI.
- Result
- DrawMe turns a single Pi prompt into a repeatable local diagram workflow with editable source, deterministic validation, exact shape lookup, and multi-format output.
Full stack
- Pi
- TypeScript
- Node.js
- npm
- draw.io
- Mermaid
Impact
- Published DrawMe as a scoped npm package with global, project-local, one-run, Git, and local checkout install paths.
- Delivered nine agent tools and three commands for planning, authoring, validating, previewing, explaining, and exporting diagrams.
- Improved diagram accuracy with structural linting, image-aware review, editable exports, and a local index of more than 10,000 official cloud and modeling shapes.
AnalyseMe Sonar Quality Analysis for Pi
Built a Pi extension that lets coding agents inspect SonarQube and SonarCloud quality data and optionally submit a fresh local scan.
- Pi
- TypeScript
- Node.js
- npm
- +2 more
Details
- Problem
- Code-quality findings often live outside the coding-agent workflow, forcing developers to switch context and manually translate quality gates, issue locations, rule guidance, and security hotspots into actionable changes.
- Approach
- Built AnalyseMe as a Pi extension that resolves project and branch context, reads SonarQube or SonarCloud quality data, returns bounded issue and hotspot details with source guidance, and optionally invokes the local SonarScanner to submit a requested fresh analysis. Credentials stay in environment variables and are masked from outputs.
- Result
- AnalyseMe brings quality-gate and finding context into Pi through read-focused tools, with one explicit scan path for validating changes against the configured Sonar server.
Full stack
- Pi
- TypeScript
- Node.js
- npm
- SonarQube
- SonarCloud
Impact
- Published AnalyseMe as a scoped npm package with global, project-local, one-run, Git, and local checkout install paths.
- Delivered six tools and two commands for quality gates, metrics, issues, source context, security hotspots, guidance, and fresh scanner submissions.
- Kept analysis reads repository-safe and masked Sonar tokens across command output, tool results, errors, and tests.
ObservMe OpenTelemetry for Pi Agent Sessions
Built a Pi extension that exports privacy-conscious traces, metrics, and logs for agent sessions to standard observability backends.
- Pi
- TypeScript
- Node.js
- npm
- +2 more
Details
- Problem
- Multi-step coding-agent sessions are difficult to troubleshoot because model calls, tools, shell activity, compaction, and child-agent work span one workflow without a shared operational view.
- Approach
- Built ObservMe as a Pi extension that maps lifecycle and agent events to OpenTelemetry traces, metrics, and logs, exports them through bounded OTLP pipelines, and propagates agent lineage plus W3C trace context across subagent processes. Content capture remains disabled by default and enabled data passes through redaction and path scrubbing.
- Result
- ObservMe makes Pi workflows observable in standard Collector and Grafana stacks without blocking sessions when telemetry is unavailable or collecting prompt and tool content by default.
Full stack
- Pi
- TypeScript
- Node.js
- npm
- OpenTelemetry
- Grafana
Impact
- Published ObservMe as a scoped npm package with global, project-local, one-run, Git, and local checkout install paths.
- Instrumented session, turn, model, tool, shell, compaction, branch, workflow, and multi-agent lineage events as OTLP traces, metrics, and logs.
- Preserved coding flow with fail-open exporters and privacy-first defaults that exclude prompts, responses, tool payloads, and shell content unless explicitly enabled.
ProtectMe Network Access Guardrails for Pi
Built a Pi extension that checks supported agent and user shell network requests against global and project allow lists before execution.
- Pi
- TypeScript
- Node.js
- npm
- +2 more
Details
- Problem
- Coding agents and direct terminal commands can contact unapproved destinations, creating accidental data exposure and dependency-download risks that are difficult to review after execution.
- Approach
- Built ProtectMe as a Pi event guard that inspects supported request-making shell commands, extracts destinations, and checks them against normalized global and trusted project allow lists. Unknown hosts fail closed, while interactive approvals can permit one request or persist a narrow host decision without storing secrets.
- Result
- ProtectMe adds transparent network-destination control to supported Pi shell flows with project-aware policy and secret-conscious audit logs, while clearly remaining a guardrail rather than a firewall or sandbox.
Full stack
- Pi
- TypeScript
- Node.js
- npm
- Network Security
- Shell
Impact
- Published ProtectMe as a scoped npm package with global, project-local, one-run, Git, and local checkout install paths.
- Guarded curl, wget, http, and https requests, including approved wrappers and options that can introduce additional destinations.
- Added default-block allow-list policy, one-time or persisted approval flows, and bounded local logs that redact common credential-bearing fragments.
BranchMe Git Workflow Automation for Pi
Built a Pi extension that verifies repository state and automates branch, worktree, integration, push, and pull request workflows.
- Pi
- TypeScript
- Node.js
- npm
- +2 more
Details
- Problem
- Coding agents need current Git context to work safely, but branch, worktree, and pull request automation can damage local history when repository state or command scope is assumed rather than verified.
- Approach
- Built BranchMe as a Pi extension that injects a bounded read-only repository snapshot and exposes strict tools for repository initialization, branch changes, linked worktrees, integration, retirement, remote updates, pushes, and GitHub pull requests. Each mutating workflow validates its repository, cleanliness, branch, and worktree preconditions before acting.
- Result
- BranchMe provides a context-aware Git workflow for Pi that supports isolated agent handoffs and pull request delivery while keeping history rewrites explicit and commit creation outside the extension.
Full stack
- Pi
- TypeScript
- Node.js
- npm
- Git
- GitHub
Impact
- Published BranchMe as a scoped npm package with global, project-local, one-run, Git, and local checkout install paths.
- Exposed 19 agent-callable tools plus a bounded read-only repository snapshot for branch, worktree, integration, push, and pull request workflows.
- Added strict safety boundaries that verify repository and worktree state, avoid force pushes, and never stage files or create user-authored commits.
CommitMe Pi Commit Assistant
Built a Pi extension that turns local git changes into safe Lightweight Conventional Commit messages and creates local commits.
- Pi
- TypeScript
- Node.js
- npm
- +2 more
Details
- Problem
- Turning a busy local diff into a clear commit message takes time, and inconsistent commit wording becomes more likely when work spans staged, unstaged, generated, or noisy files.
- Approach
- Built CommitMe as a Pi extension and npm package that reads local git status, staged and unstaged paths, diff stats, redacted diff excerpts, and safe project metadata. It builds a compact prompt for the active Pi model, supports optional steering text and confirmation, validates the Lightweight Conventional Commit shape, then stages changes and creates a local commit without pushing.
- Result
- CommitMe provides a repeatable local workflow for drafting, reviewing, and creating clear commits from actual repository changes while keeping safety checks, redaction, and no-telemetry behavior built in.
Full stack
- Pi
- TypeScript
- Node.js
- npm
- Git
- LLM Tooling
Impact
- Published CommitMe as a scoped npm package with slash-command, agent-tool, global, project-local, one-run, Git, and local checkout install paths.
- Improved commit quality by gathering staged and unstaged git context, trimming noisy diffs, accepting steering text, and producing Lightweight Conventional Commit messages.
- Added safety guardrails that redact sensitive context, refuse known secret files and high-confidence tokens, recheck repository status before staging, and never push.
GuardMe Pi Tool-Call Safety Guardrails
Built a Pi extension that enforces deny-first policy for shell and filesystem tool calls with YAML rules, path protections, and approval flows.
- Pi
- TypeScript
- Node.js
- npm
- +3 more
Details
- Problem
- Coding agents can call powerful local shell and filesystem tools with full user permissions, so a mistaken or overly broad model request can read secrets, mutate protected files, or run risky commands before the user notices.
- Approach
- Built GuardMe as a Pi extension that checks shell and file tool calls before execution. It merges global and project YAML policy, evaluates compound shell commands by executable segment, protects sensitive paths and generated shell-like content, and uses in-session approval only after first blocking and coaching risky actions that are not hard-denied.
- Result
- GuardMe gives Pi sessions IAM-like deny-first guardrails while staying configurable through /guardme and policy files, making local AI-assisted development safer without presenting itself as an OS sandbox.
Full stack
- Pi
- TypeScript
- Node.js
- npm
- YAML
- Security
- LLM Tooling
Impact
- Published GuardMe as a scoped npm package with global, project-local, one-run, Git, and local checkout install paths for Pi.
- Reduced accidental risky model actions with deny-first shell and filesystem guards for cloud CLIs, credentials, `.git`, `.env`, destructive commands, and sensitive paths.
- Added a policy-driven approval workflow that merges global and project YAML rules, tracks warned fingerprints, and supports safe one-time or persisted decisions.
IssueMe GitHub Issue Management for Pi
Built a Pi extension that gives coding agents repository-scoped GitHub issue workflows through structured API tools and bounded local context.
- Pi
- TypeScript
- Node.js
- npm
- +2 more
Details
- Problem
- Agents need complete issue context to plan and track work, but raw GitHub API usage creates authentication, repository-boundary, pagination, destructive-action, and oversized-response risks.
- Approach
- Built IssueMe as a Pi extension that resolves the active repository, communicates with GitHub through REST and GraphQL APIs, stores bounded non-secret issue context locally, and exposes structured tools for issue lifecycle, planning metadata, Projects v2, sub-issues, and linked development.
- Result
- IssueMe provides an agent-friendly issue management layer that supports discovery and updates from Pi while keeping repository scope, local cache behavior, token handling, and destructive confirmations explicit.
Full stack
- Pi
- TypeScript
- Node.js
- npm
- GitHub REST API
- GraphQL
Impact
- Published IssueMe as a scoped npm package with global, project-local, one-run, Git, and local checkout install paths.
- Exposed 29 tools for issues, labels, milestones, comments, assignees, Projects v2, sub-issues, development links, deletion, and bulk workflows.
- Kept agent context bounded with a local issue cache while protecting tokens and requiring explicit confirmation for destructive taxonomy and deletion operations.
MicMe Local Voice-to-Text Pi Extension
Built a local-first voice-to-text extension for Pi that records coding prompts, transcribes locally, and inserts reviewed transcripts into the editor.
- Pi
- TypeScript
- Node.js
- npm
- +2 more
Details
- Problem
- Writing longer coding prompts in a terminal can interrupt flow, while cloud dictation tools add privacy concerns and extra moving parts.
- Approach
- Built MicMe as a Pi extension that records microphone input with ffmpeg, transcribes locally through whisper.cpp, Python Whisper, or a custom backend, and pastes the transcript into Pi for review before submission. Added configuration, device discovery, diagnostics, and npm publishing workflows so the package is practical to install and maintain.
- Result
- MicMe provides a local-first voice input workflow for coding prompts, letting users speak, review, and submit transcripts without leaving the Pi editor.
Full stack
- Pi
- TypeScript
- Node.js
- npm
- ffmpeg
- whisper.cpp
Impact
- Published MicMe as a scoped npm package with global, project-local, one-run, Git, and local checkout install paths.
- Preserved privacy by keeping speech-to-text local by default with ffmpeg, whisper.cpp, Python Whisper, or custom backends.
- Improved coding-agent prompt flow with toggle recording, review-first transcript insertion, device discovery, configuration, and diagnostics.
Pi Code DevOps Automation Workspace
Built a project-local Pi workspace that combines agent orchestration, safety guardrails, and reusable skills for reliable automation.
- Pi
- TypeScript
- Node.js
- AWS
- +1 more
Details
- Problem
- Ad hoc automation workflows are difficult to scale safely when every task uses different prompts, tools, and execution patterns.
- Approach
- Created a DevOps-focused Pi workspace with custom extensions, agent teams, chain pipelines, and reusable skills. Added guardrails for filesystem writes and outbound hosts, plus utility tools for AWS docs lookup, concise summarization, commit generation, and token usage reporting.
- Result
- The workspace enables safer, repeatable, and auditable engineering workflows across planning, coding, QA, and documentation.
Full stack
- Pi
- TypeScript
- Node.js
- AWS
- YAML
Impact
- Standardized multi-agent workflows for planning, implementation, QA, and documentation.
- Improved operational safety with write guards, destructive-command blocking, and host allowlists.
- Increased delivery speed with reusable chains, specialist agents, and deterministic utility tools.
Terraform Skills Bundle for Codex CLI
Created a reusable Codex CLI skill bundle to standardize Terraform planning, module delivery, review, and AWS reporting.
- Terraform
- AWS CLI
- Bash
- MCP
- +2 more
Details
- Problem
- Terraform delivery quality often varies when teams rely on inconsistent templates, review depth, and planning habits.
- Approach
- Packaged opinionated Codex skills, scripts, and templates for memory bootstrap, Terraform planning, module creation, code review, and AWS account reporting, with MCP-aware context support.
- Result
- The bundle enables faster, more predictable Terraform and AWS delivery with shared standards, stronger validation gates, and less rework.
Full stack
- Terraform
- AWS CLI
- Bash
- MCP
- Codex CLI
- LocalStack
Impact
- Reduced repeated setup work with memory-bank bootstrap and reusable project rules.
- Improved consistency through guided workflows for child modules, root modules, planning, and review.
- Added AWS reporting coverage for cost, IAM, TLS, and network inventory.
Calm&Clear Serverless AWS Platform
Built a secure multi-account serverless backend that reduced operational overhead and improved integration reliability.
- Terraform
- AWS Organizations
- CloudFront
- API Gateway
- +8 more
Details
- Problem
- The client needed a secure and scalable integration backend without adding long-term infrastructure overhead.
- Approach
- Implemented a multi-account AWS foundation and built a serverless integration layer on API Gateway, Lambda, and DynamoDB. Secured public entry points with WAF, custom domains, and Route 53 routing, then added centralized observability for API and function behavior.
- Result
- The platform now supports faster delivery, clearer environment boundaries, and a scalable integration architecture ready for additional business workflows.
Full stack
- Terraform
- AWS Organizations
- CloudFront
- API Gateway
- Lambda
- DynamoDB
- WAF
- ECS
- EC2
- RDS
- S3
- SES
Impact
- Established isolated development, staging, and production accounts to strengthen governance and release safety.
- Replaced server-managed integration flows with API Gateway, Lambda, and DynamoDB to lower maintenance effort.
- Improved payment and invoicing reliability with centralized API monitoring, logging, and tracing.
NeuroEAye Secure AWS Foundation
Delivered secure, automated AWS infrastructure for healthcare workloads with repeatable CI/CD and lower release risk.
- AWS Organizations
- AWS SecurityHub
- Terraform
- ECS
- +8 more
Details
- Problem
- Healthcare-related applications required a secure and repeatable cloud foundation where reliability and controlled delivery were critical.
- Approach
- Automated AWS infrastructure provisioning with Terraform, containerized workloads with Docker, and implemented CI/CD pipelines with GitHub Actions and AWS CodePipeline.
- Result
- The project gained safer release workflows, consistent infrastructure delivery, and a cloud baseline ready for regulated growth.
Full stack
- AWS Organizations
- AWS SecurityHub
- Terraform
- ECS
- CloudFront
- EFS
- RDS
- S3
- Docker
- CloudMap
- GitHub Actions
- CodePipeline
Impact
- Reduced configuration drift by standardizing infrastructure provisioning with Terraform.
- Improved release consistency through Dockerized workloads and automated pipelines.
- Strengthened security posture across infrastructure and delivery workflows for healthcare use cases.
S3 Explorer Browser File Manager
Built a lightweight browser-only S3 manager with session-only credentials and no backend dependency.
- HTML
- CSS
- Amazon S3
- AWS SDK for JavaScript v2
Details
- Problem
- Managing S3 objects often requires heavyweight tooling, while ad hoc scripts can be difficult for non-CLI users.
- Approach
- Implemented a browser-native S3 utility using AWS SDK v2 from a CDN, then expanded UX with breadcrumbs, filtering, previews, multipart uploads, bulk actions, and theme switching.
- Result
- The app delivers practical day-to-day S3 operations in a minimal interface with improved code structure and no server-side credential storage.
Full stack
- HTML
- CSS
- Amazon S3
- AWS SDK for JavaScript v2
Impact
- Enabled direct S3 workflows in the browser, including browse, upload, download, move, and delete.
- Improved maintainability by refactoring a single-file app into modular CSS and JavaScript assets.
- Preserved security posture by keeping credentials in memory only and never persisting secrets.
Devbox Global Machine Bootstrap
Automates fresh macOS and Linux workstation setup for DevOps tooling through a reusable shell bootstrap.
- Bash
- Linux
- macOS
- Git
Details
- Problem
- Setting up a new development machine repeatedly is slow and error-prone, especially across different operating systems.
- Approach
- Built a function-based shell bootstrap project that detects the host OS, installs required software, logs progress, and handles failures consistently.
- Result
- The project provides a repeatable, low-friction way to prepare DevOps workstations while reducing configuration drift between machines.
Full stack
- Bash
- Linux
- macOS
- Git
Impact
- Reduced manual onboarding effort by automating baseline machine setup steps.
- Standardized workstation provisioning across macOS and Linux with one entrypoint.
- Improved setup reliability through explicit OS detection, logging, and error handling.
UK Home Office DSP Serverless Delivery
Contributed to high-assurance serverless AWS delivery, improving automation, observability, and governance in a public-sector program.
- AWS Lambda
- API Gateway
- CloudFront
- RDS
- +4 more
Details
- Problem
- A large public-sector digital service required resilient cloud infrastructure with strong security controls and dependable deployment automation.
- Approach
- Worked in a global team to implement serverless AWS architecture using API Gateway, Lambda, Aurora, S3, and CloudFront. Strengthened CI/CD with GitHub Actions and expanded monitoring through CloudWatch and Grafana.
- Result
- The platform achieved more reliable releases, stronger governance controls, and clearer operational visibility across distributed teams.
Full stack
- AWS Lambda
- API Gateway
- CloudFront
- RDS
- S3
- CloudWatch
- GitHub Actions
- Grafana
Impact
- Accelerated release flow by automating quality checks, security scans, and deployments.
- Improved operational transparency through CloudWatch and Grafana dashboards.
- Supported secure, high-availability architecture patterns for critical public services.
Personal Observability Engineering Lab
Built a personal observability sandbox to validate logs, metrics, and tracing patterns before production rollout.
- Grafana
- Loki
- Prometheus
- Tempo
- +2 more
Details
- Problem
- Experimenting directly in production observability environments is costly and risky, especially when testing alerting and correlation changes.
- Approach
- Built a local lab stack with Grafana, Loki, Prometheus, Tempo, and OpenTelemetry to prototype dashboards, alerts, and trace-to-log workflows under repeatable test conditions.
- Result
- The lab serves as a reliable proving ground for observability decisions, improving confidence and reducing risk before production rollout.
Full stack
- Grafana
- Loki
- Prometheus
- Tempo
- OpenTelemetry
- Docker
Impact
- Tested dashboard and alerting patterns in a safe environment before applying them to client systems.
- Improved troubleshooting workflows through end-to-end log, metric, and trace correlation.
- Reduced rollout risk by validating runbooks against realistic telemetry scenarios.
Docker Compose Remote Control API
Developed a Dockerized Flask API behind Gunicorn and Nginx for authenticated remote docker-compose operations.
- Python
- Flask
- Docker Compose
- Gunicorn
- +2 more
Details
- Problem
- Operating occasional infrastructure services on remote hosts is inconvenient when shell access is the only control path.
- Approach
- Built a REST API with login and command endpoints, constrained execution to docker-compose commands, and packaged the service as Flask + Gunicorn behind Nginx.
- Result
- Operators can trigger compose workflows from external tools (for example Grafana buttons) without exposing full interactive shell access.
Full stack
- Python
- Flask
- Docker Compose
- Gunicorn
- Nginx
- JWT
Impact
- Enabled secure remote control of auxiliary container services through REST endpoints.
- Protected command execution with JWT-based authentication and authorization.
- Simplified operations with containerized deployment and clear API/Gunicorn/Nginx log paths.
Typla.nl AWS Reliability & Support Program
Improved stability, security, and delivery predictability across multi-client AWS environments through IaC and pipeline standardization.
- Terraform
- AWS Organizations
- OpenTelemetry
- Lambda
- +13 more
Details
- Problem
- A mixed client portfolio running on EC2 and ECS created operational complexity around reliability, governance, and consistent delivery.
- Approach
- Focused on infrastructure hardening and optimization across AWS accounts while improving release workflows with CodePipeline, Jenkins, Ansible, and GitHub-based version control. Standardized infrastructure change management through Terraform.
- Result
- The environment became easier to operate and scale, with clearer account boundaries, stronger delivery consistency, and lower operational risk.
Full stack
- Terraform
- AWS Organizations
- OpenTelemetry
- Lambda
- AWS EC2
- AWS ECS
- Cognito
- S3
- CloudFront
- SES
- SNS
- SQS
- WAF
- CloudMap
- CodePipeline
- Jenkins
- Ansible
Impact
- Strengthened governance and security across shared multi-account AWS environments.
- Improved deployment consistency across client applications with standardized CI/CD workflows.
- Reduced infrastructure drift by managing platform changes through Terraform.
Lundbeck Batch Tracking AWS Launch
Delivered production AWS infrastructure under launch deadlines, combining ECS reliability with CloudFront performance gains.
- AWS ECS
- AWS CloudFront
- AWS CodePipeline
- Terraform
Details
- Problem
- A newly developed batch-tracking application required secure and dependable AWS infrastructure before release.
- Approach
- Implemented an ECS-centered backend architecture and configured CloudFront for frontend delivery performance, using infrastructure patterns focused on reliability, maintainability, and predictable operations.
- Result
- The application launched on a stable cloud foundation with consistent backend behavior and improved end-user delivery performance.
Full stack
- AWS ECS
- AWS CloudFront
- AWS CodePipeline
- Terraform
Impact
- Delivered a production-ready cloud environment aligned with launch milestones.
- Improved frontend delivery performance through CloudFront edge distribution.
- Ensured stable backend operations with an ECS-based architecture.
Construction Site Inventory AWS Launch Foundation
Delivered a secure, cost-aware AWS launch baseline with private asset delivery and resilient container services.
- AWS ECS
- AWS S3
- AWS CloudFront
- Terraform
- +1 more
Details
- Problem
- The client needed a practical AWS baseline for a new product with clear environment separation and controlled access to stored assets.
- Approach
- Provisioned development and production infrastructure in a new AWS account, delivered frontend assets through CloudFront with private S3 origins, and deployed backend services on ECS using high-availability design principles.
- Result
- The team received a stable and maintainable cloud setup that balanced security, cost, and scalability for early product growth.
Full stack
- AWS ECS
- AWS S3
- AWS CloudFront
- Terraform
- IAM
Impact
- Provisioned separate development and production environments from day one.
- Secured static asset delivery using private S3 origins behind CloudFront.
- Delivered an ECS backend foundation aligned with reliability and budget constraints.
YouChooseWork AWS Launch Infrastructure
Designed and delivered launch-ready AWS infrastructure balancing performance, reliability, and long-term maintainability.
- AWS ECS
- AWS CloudFront
- Terraform
- AWS CloudWatch
- +1 more
Details
- Problem
- The platform needed dependable AWS infrastructure to support launch goals while maintaining performance and operational stability.
- Approach
- Designed and provisioned AWS resources for backend and frontend delivery, using ECS for service resilience and CloudFront for low-latency content distribution.
- Result
- The product launched on a cloud setup that balanced reliability, performance, and maintainability for ongoing growth.
Full stack
- AWS ECS
- AWS CloudFront
- Terraform
- AWS CloudWatch
- SNS
Impact
- Delivered a production-ready backend environment on a highly available ECS architecture.
- Improved user-facing performance with CloudFront distribution.
- Accelerated launch readiness with repeatable infrastructure provisioning patterns.
Compli AWS Observability Baseline
Unified AWS monitoring into a shared observability baseline that improved alert quality and sped up incident triage.
- AWS CloudWatch
- Grafana
- InfluxDB
- Telegraf
- +2 more
Details
- Problem
- Monitoring was fragmented across services, slowing incident triage and reducing confidence in alerts.
- Approach
- Designed a centralized observability layer with CloudWatch as the primary source and Grafana as the shared visualization surface. Added InfluxDB and Telegraf for deeper telemetry coverage, then connected alerting to Slack support channels for faster team response.
- Result
- Teams moved to a consistent monitoring model with clearer alert ownership, less noisy escalation, and a faster path from detection to resolution.
Full stack
- AWS CloudWatch
- Grafana
- InfluxDB
- Telegraf
- Slack
- SNS
Impact
- Centralized CloudWatch telemetry into shared Grafana dashboards used across workloads.
- Introduced Slack-routed alert workflows with clearer ownership and faster first response.
- Extended diagnostic depth by combining CloudWatch with InfluxDB and Telegraf metrics.
Nebula Internal EKS Platform Enablement
Helped standardize internal EKS platform provisioning, improving environment consistency and onboarding speed for teams.
- AWS EKS
- Terraform
- Kubernetes
- AWS Networking
- +2 more
Details
- Problem
- Internal teams needed a dependable AWS platform baseline to avoid repeated manual setup and inconsistent environment quality.
- Approach
- Contributed to an internal EKS platform initiative that provisions core infrastructure with Terraform, with emphasis on repeatability, resilient networking, and operational consistency.
- Result
- Teams gained a predictable platform foundation that shortened project startup time and improved long-term maintainability.
Full stack
- AWS EKS
- Terraform
- Kubernetes
- AWS Networking
- GitHub Actions
- OpenVPN
Impact
- Standardized baseline infrastructure patterns for new internal development projects.
- Improved consistency and repeatability through Terraform-driven provisioning.
- Reduced setup time for teams launching new platform workloads.
Office Reservation Manager AWS Delivery Baseline
Automated Mendix environment provisioning and CI workflows to reduce setup effort and improve release consistency.
- AWS CloudFormation
- GitHub Actions
- Mendix
- ECS
- +2 more
Details
- Problem
- The team needed a dependable development environment for a Mendix application with minimal setup friction and repeatable delivery steps.
- Approach
- Implemented infrastructure provisioning through AWS CloudFormation and standardized CI pipeline execution with GitHub Actions.
- Result
- Environment setup became faster and more reliable, enabling smoother day-to-day development and release preparation.
Full stack
- AWS CloudFormation
- GitHub Actions
- Mendix
- ECS
- EFS
- RDS
Impact
- Delivered a ready-to-use development environment for application teams.
- Reduced manual provisioning work through CloudFormation automation.
- Improved CI reliability with repeatable GitHub Actions workflows.
No projects match this filter yet.
Experience
Experience includes full-time DevOps progression at Valcon, freelance AWS delivery through Upwork, and earlier operations leadership at INA Grupa.
DevOps Engineer — Principal Technology CSE
Valcon (full-time)
Jan 2026 — Present
- Focus: Amazon Web Services (AWS), Terraform, and platform engineering.
- Lead enterprise cloud delivery with secure and scalable infrastructure patterns.
DevOps Engineer
Upwork (freelance)
Feb 2025 — Present
- Focus: Amazon Web Services (AWS) and Terraform.
- Design and operate production-ready AWS environments for freelance clients.
DevOps Engineer — Senior Specialist
Valcon (full-time)
Jul 2024 — Jan 2026
- Focus: Amazon Web Services (AWS) and Terraform.
- Scaled reusable IaC and deployment workflows across client engagements.
DevOps Technology Specialist
Valcon (full-time)
Jan 2023 — Jul 2024
- Focus: infrastructure and Linux.
- Supported stable operations and core infrastructure improvements.
DevOps Engineer
Valcon (full-time)
Sep 2022 — Jan 2023
- Focus: infrastructure, Git, and DevOps tooling.
- Built foundational delivery workflows for cloud platform work.
Technical Support Manager
INA Grupa (full-time)
Jan 2011 — Nov 2022
- Led long-term on-site production support operations.
- Built incident communication and coordination discipline applied in later DevOps delivery.
Skills
Production-focused tooling used across full-time and freelance cloud delivery projects.
AWS Platform Delivery
Tools & Automation
Frameworks and Languages Used Across Projects
Observability & Security
LLM Tooling
Certifications
AWS Certified Solutions Architect — Associate
Amazon Web Services
Verify credential
About
I work in two delivery modes: embedded partner for hiring teams and focused consultant for freelance clients. In both cases, the goal is the same simple architecture first, reliable automation, strong security posture, and clear operational ownership after handover.
For hiring managers
Dependable ownership from platform design to stable day-2 operations.
For engineering teams
Practical CI/CD and infrastructure automation that removes bottlenecks without overengineering.
For freelance clients
Security-first AWS delivery focused on measurable business outcomes and long-term maintainability.
Contact
Freelance clients
Start with a focused infrastructure review: top risks, practical next steps, and a clear scope and handover recommendation.