AI Internal Tools

About

Andrei Alexandru Gabriel

AI automation & applied engineering for real operations

Bucharest, Romania

Engineer with 10+ years in military aircraft maintenance and quality leadership, PhD work in predictive maintenance AI. I build automations, internal tools, and LLM features that teams actually use—not slide decks.

Workflow automation, agents, knowledge assistants, and dashboards. This site hosts demos; contact is for production scoping.

Hiring: full-stack delivery on React / Next.js / TypeScript, Supabase, Vercel, plus applied AI and automation—end to end.

Request automationContact

Selected roles

  • Applied AI engineer
  • Automation & integrations
  • QA & operational systems

Stack

Core application stack

Typical delivery setup (including products like PixelPlot on Vercel). Patterns and integrations matter more than listing every model name.

  • Frontend

    React / Next.js / TypeScript

    • HTML / CSS
    • JavaScript
    • Component-driven UI

  • Backend

    Supabase (Postgres, Auth, Storage)

    • Edge / server functions when needed
    • Realtime when applicable

  • Deployment

    Vercel

    • Environments & deploy workflows
    • Domains & config

  • Applied AI

    LLM integrations, multimodal workflows, structured AI pipelines

    • OpenAI & Anthropic APIs
    • Prompt orchestration
    • Structured outputs & tool calling

  • AI product patterns

    Shippable AI features—not chasing model names alone

    • Prompt pipelines & generation workflows
    • Async jobs, queues, API chaining
    • Media / multimodal flows where relevant

  • Automation

    n8n, API orchestration, workflow automation

    • n8n: multi-step workflows, webhooks, SaaS integrations
    • API orchestration between tools and internal services

  • Agents & retrieval

    Agents, RAG, MCP servers, Google ADK

    • MCP (Model Context Protocol): servers, tools, client integrations
    • Google ADK (Agent Development Kit) for agent design and delivery
    • RAG, retrieval workflows & agent orchestration

Recognition

  • PixelPlot

    2024 — present

    Co-founder · full-stack (AI focus)

    PixelPlot is an AI creative tool for on-brand short-form ads and storyboards: brand and product context in, then generated assets for social (e.g. Meta, TikTok, LinkedIn). Co-founder: full-stack engineering with an AI focus—app, integrations, LLM and multimodal generation workflows.

    React, Next.js, TypeScript, Supabase, Vercel, OpenAI, image & video-style generation

  • Romanian Air Force — Quality assurance

    2017 — present

    Head of QA

    Leads QA in aircraft maintenance—audits, workflows, data for leadership, high-reliability context.

    QMS, maintenance records

  • ICI Bucharest

    2024 — 2025

    Blockchain product (part-time)

    Decentralized maintenance data tracking—research to product, partner presentations.

    DApps, smart contracts

Earlier: MiG-21 engine/airframes lead, production & planning (2014–2017).

Education

  • PhD (summa cum laude)

    2019 — 2022

    Politehnica Bucharest — Industrial & robotic engineering

    Predictive maintenance QA, ML, blockchain-oriented records.

  • AI & GPT Bootcamp

    2024

    Encode Club

    Agents, RAG, multimodal. Project: aviation RAG assistant.

    Encode bootcamp project Project repo →

  • Solidity Bootcamp

    2023

    Encode Club

    Smart contracts, DApps, EVM security.

  • MSc & BSc — Aerospace engineering

    2010 — 2016

    Military Technical Academy

    Aerospace engineering & maintenance.

Publications

Peer-reviewed papers and articles.

Also

Research, domain, and chain work beyond the core web + AI stack above.

  • Research & ML

    Traditional ML & forecasting · Python · Data storytelling

  • Domain

    Aviation maintenance & QA leadership · Predictive maintenance

  • Other

    Solidity / EVM · Blockchain product work

Languages

  • Romanian — Native
  • English — Advanced
  • Italian — Beginner

Let's talk

Describe a workflow; I'll suggest what to automate first.

ContactDemosHome