Mircea
Stefanescu

REMOTE FREELANCE DEVELOPMENT

Software you
can inspect.

I’m Mircea. I build APIs, web applications and the workflows around them. Explore the code, see the demos and find the starting point for your next project.

Java & Spring Boot · Web applications · Delivery automation

EUR 40/hour gross · Optional EUR 200 diagnostic for one agreed issue. See scope and terms.

16project demos
8technical areas
100%remote collaboration
Angular + Spring Boot
Registration and product catalog in the local Angular and Spring Boot application
DEMO SNAPSHOT A connected frontend and Java backend.Explore this project

START WITH ONE CONCRETE PROBLEM

A useful first step,
with a clear scope.

For SaaS teams, software agencies and independent product teams: choose the issue you need to understand, then agree a bounded diagnostic or implementation.

An API or integration behaves unpredictably

Reproduce one failure, trace the contract and retry behavior, and define a correction with acceptance checks.

API and integration diagnosis ↗

A pipeline blocks reliable releases

Review one build or deployment path, identify failure and recovery gaps, and agree the next repair.

CI/CD pipeline diagnosis ↗

A prompt change needs repeatable checks

Define representative cases, expected outputs and a report that exposes regressions and unscored results.

AI regression review ↗

EUR 200 gross diagnostic: one agreed issue, up to five work hours, with a written findings report and recommendations within seven calendar days after receipt of the agreed inputs. Implementation and ongoing support are scoped separately. Hourly work is EUR 40 gross, before platform commissions. Scope, payment terms and availability are agreed before work starts.

These project examples are personal demonstrations, not customer case studies. Remote freelance/B2B work from Romania; access and delivery arrangements are agreed per engagement.

THE PORTFOLIO

From an API to the
systems around it.

Browse by the problem you’re working on. Every project links to its source and setup instructions. Captures show local demonstrations; each project explains the scope of that evidence.

01

Web applications and SaaS

Build a customer workflow, dashboard or internal tool with a connected backend.

Registration and product catalog in the running local web application.

PROJECT 06

Web applications with a Java backend

Registration, catalog and account workflows, frontend/API integration, cart and order views, tests and setup documentation.

Java, Spring Boot, Angular, TypeScript, SQL, JWT

Demo context

The demonstrated checkout uses simulated payments. A real payment provider requires separate configuration and acceptance checks.

Explore source & setup : Web applications with a Java backend
Check-in and streak workflow in the running local SaaS demonstration.

PROJECT 08

SaaS dashboards and recurring workflows

Account workflows, time-zone-aware schedules, recurring tasks, streaks, groups, exports and background-job integration.

React, Express, PostgreSQL, Redis, BullMQ

Demo context

Identity, subscription and notification integrations use demonstrations. No real subscriptions are charged.

Explore source & setup : SaaS dashboards and recurring workflows
02

APIs and integrations

Build or repair APIs, connect services and handle asynchronous business workflows.

Recovery evidence from the local commerce services.

PROJECT 03

Persistent APIs and resilient order workflows

Authenticated endpoints, database persistence, duplicate-safe workflows, broker recovery and integration tests.

Python, FastAPI, PostgreSQL, Kafka, RabbitMQ, Kubernetes

Demo context

Payment and refund operations use a local simulation. Live payment-provider integration must be scoped and validated separately.

Explore source & setup : Persistent APIs and resilient order workflows
Successful order saga in the local event-driven demonstration.

PROJECT 05

Java event-driven integrations

Kafka integrations, event contracts, order and inventory APIs, idempotent processing, saga compensation and tests.

Java, Spring Boot, Kafka, Schema Registry, H2

Demo context

Payment, shipping and notification providers are simulated. Customer authentication and live providers are separate integration work.

Explore source & setup : Java event-driven integrations
03

Real-time collaboration

Add live messaging, presence and persistent collaboration history.

Bidirectional messages between two local demonstration sessions.

PROJECT 14

Real-time collaboration features

Rooms, bidirectional messaging, typing and presence indicators, rate limits, persistent history and integration tests.

Java, Spring Boot, WebSocket/STOMP, Angular, Kafka, PostgreSQL, Redis

Demo context

The demonstration uses local identities. Your external OAuth provider and production access policies need separate configuration.

Explore source & setup : Real-time collaboration features
04

DevOps and GitOps

Automate delivery, infrastructure templates and repeatable Kubernetes releases.

Generated pipeline configuration in the demonstration portal.

PROJECT 01

Developer self-service portals

A portal with access roles, pipeline and infrastructure templates, health views and documented deployment handoff.

JavaScript, Node.js, Express, OIDC/RBAC, Jenkins, Terraform, Helm

Demo context

The portfolio uses local integrations and templates. Live cloud provisioning and your identity provider require separate configuration.

Explore source & setup : Developer self-service portals
Deployment metadata captured from a local release workflow.

PROJECT 02

Build, test and deployment automation

A reproducible build/test pipeline, container delivery, release checks and a documented rolling, blue/green or canary deployment workflow.

Jenkins, Docker, Terraform, Ansible, Kubernetes, Helm

Demo context

Local deployment and recovery demonstrations do not establish a production uptime guarantee.

Explore source & setup : Build, test and deployment automation
Delivery metrics dashboard populated by the local demonstration ledger.

PROJECT 07

Git-based Kubernetes delivery and recovery

Argo CD reconciliation, Helm environments, promotion and rollback workflows, backup/restore checks and delivery dashboards.

Kubernetes, Argo CD, Helm, Terraform, Velero, Prometheus, Grafana

Demo context

The deployment examples are local and the DORA ledger is synthetic. It does not measure a customer's delivery performance.

Explore source & setup : Git-based Kubernetes delivery and recovery
05

Security and secrets

Add supply chain checks, signed artifacts and controlled secrets access.

SBOM inventory generated for the demonstration artifact.

PROJECT 04

Software supply chain checks

SBOM generation, vulnerability and license checks, signed artifacts and provenance, tamper-rejection tests and release policies.

CycloneDX, Trivy, Grype, Cosign, Jenkins, policy checks

Demo context

These are implemented pipeline controls, not an independent security audit or compliance certification.

Explore source & setup : Software supply chain checks
Operational dashboard for the local secrets toolkit.

PROJECT 16

Secrets integration and operational visibility

Vault AppRole integration, access policies, rotation, audit logs, dashboards and tested recovery procedures.

Vault, Terraform, Node.js, Elasticsearch, Prometheus, Grafana

Demo context

KMS and regional failover are local demonstrations. No live cloud certification or compliance assurance is claimed.

Explore source & setup : Secrets integration and operational visibility
06

Observability and FinOps

Make service behavior and resource allocation easier to understand.

Grafana dashboard connected to local demonstration telemetry.

PROJECT 11

Metrics, logs, traces and operational dashboards

Instrumentation, dashboards, alert routing, retry/deduplication checks and operational troubleshooting notes.

OpenTelemetry, Prometheus, Grafana, Mimir, logs and traces

Demo context

Alert delivery uses local receivers. An external New Relic account or live paging provider is not demonstrated.

Explore source & setup : Metrics, logs, traces and operational dashboards
Cost allocation report generated from demonstration data.

PROJECT 09

Kubernetes resource and cost visibility

Allocation reports, budget and anomaly checks, right-sizing recommendations and guarded idle-workload stop/restore workflows.

Kubernetes, OpenCost/Kubecost, Python, HPA, cost reporting

Demo context

Inputs and estimates are synthetic/local. No measured customer savings or live cloud bill is claimed.

Explore source & setup : Kubernetes resource and cost visibility
07

Streaming data and analytics

Process events into persistent results and queryable analytics.

Prediction results returned by the local analytics API.

PROJECT 12

Streaming pipelines and analytics APIs

Kafka/Flink processing, persistence, scoring integration, result APIs, checkpoint recovery and relevant tests.

Java, Apache Kafka, Apache Flink, Spring Boot, PostgreSQL, Elasticsearch, KEDA

Demo context

The demonstration uses synthetic sensor events. It does not establish prediction accuracy on a customer's data.

Explore source & setup : Streaming pipelines and analytics APIs
08

AI integration and evaluation

Integrate document retrieval and add repeatable checks for AI features.

Successful query against the local knowledge API.

PROJECT 15

Document retrieval and RAG APIs

Authenticated ingestion/query endpoints, role checks, hybrid retrieval, caching, evaluation and vector-store recovery.

Java, Spring Boot, JWT/RBAC, Qdrant, hybrid retrieval, Kubernetes

Demo context

Answer generation uses a local mock. Real-model quality and a customer's document collection require separate evaluation.

Explore source & setup : Document retrieval and RAG APIs
Evaluation summary from an executed local run.

PROJECT 10

Repeatable evaluation for AI features

Versioned evaluation datasets, output and tool-call checks, cached runs, persisted reports and CI integration.

Python, SQLite, evaluation datasets, HTTP adapters, CI

Demo context

Controlled mock/HTTP fixtures test evaluation and integration behavior. Your target model and datasets need their own evaluation.

Explore source & setup : Repeatable evaluation for AI features
Prompt evaluation results from an executed local run.

PROJECT 13

Prompt regression and response checks

Versioned prompt cases, output/reference checks, injection test cases, provider-adapter tests and repeatable reports.

Python, structured outputs, prompt evaluation, test automation

Demo context

Passing the supplied cases does not guarantee safety or quality for arbitrary prompts or models.

Explore source & setup : Prompt regression and response checks

HOW I WORK

Clear scope.
Visible progress.
A usable handoff.

I work remotely on defined software projects, from focused API fixes to connected applications and delivery workflows.

  1. Start with the actual problem.Agree the outcome, inputs and acceptance criteria before implementation.
  2. Make the work inspectable.Use source code, relevant tests and demonstrations to review progress.
  3. Leave a useful handoff.Document setup, configuration and the decisions needed to maintain the result.
Read the complete service catalog

LET’S TALK ABOUT YOUR PROJECT

What needs to
work better?

Describe the current behavior, your stack and the outcome you need. We can agree the scope, acceptance checks and availability from there.

Open the project enquiry form

If useful, include reference GITHUB-PORTFOLIO in your enquiry. Share the problem and stack, without passwords or production secrets.