Skip to content

Software Engineer — Cloud & AI Engineering · AWS

Matteo Carola

I build cloud infrastructure and Generative AI platforms on AWS: Infrastructure as Code, serverless architectures, CI/CD, LLM integration (RAG, prompt engineering) and hands-on operation of production environments.

Naples, Italy

Email me GitHub LinkedIn

My personal assistant

A RAG chatbot indexed on everything I’ve built — including itself. Ask it about my projects, the stack I use or my path.

About

I’m a Software Engineer with two years of experience on enterprise and public-sector projects, having grown from web application development to building cloud and Artificial Intelligence solutions on AWS.

I personally implemented the AWS infrastructure of DELFI, a Generative AI platform for the analysis of judicial documents, working closely with the Cloud Lead; and I delivered, largely on my own, the AI & Digital Twin lab of an ITS foundation funded by the PNRR programme, from systems setup all the way to acceptance testing.

I use AI-assisted development tools in my daily workflow. In parallel I’m completing my bachelor’s degree in Computer Engineering and preparing for the AWS Certified AI Practitioner (AIF-C01) certification.

Skills

Cloud & DevOps

  • AWS (Lambda, Step Functions, API Gateway, S3, DynamoDB, EC2/GPU, CloudFront, Cognito, OpenSearch, CloudWatch, CloudTrail, IAM, SSM)
  • Infrastructure as Code (CloudFormation)
  • CI/CD (Bitbucket Pipelines, OIDC)
  • Linux (Ubuntu Server)
  • Kubernetes (k3s)
  • Tailscale

AI & Generative AI

  • Open-source LLMs (Mistral, Ollama)
  • RAG and semantic search (OpenSearch, embeddings)
  • Prompt engineering
  • MLflow
  • JupyterHub
  • scikit-learn (fundamentals)
  • AWS Bedrock (architectural evaluation)
  • AI-assisted development (coding agents)

Development

  • TypeScript
  • JavaScript
  • React
  • Python (Lambda, scripting, FastAPI)
  • REST APIs
  • Git
  • Bitbucket
  • GitHub

Methodology & Delivery

  • Agile / Scrum
  • Jira
  • Operational project management
  • Incident management and root cause analysis
  • Technical, acceptance and DNSH-compliance documentation (public sector / PNRR)
Amazon RDS per Oracle supporta Oracle Database 26ai con Bedrock — 9 Jul 2026 Amazon S3 Vectors per AI e RAG — 9 Jul 2026 Automazione Migrazione Database con AI in AWS DMS — 13 Jul 2026 Colibri: LLM GLM-5.2 (744B MoE) su macchine consumer — 13 Jul 2026 Troubleshooting Spark su EMR on EKS con AI — 13 Jul 2026 Istanze Amazon EC2 G7e con GPU NVIDIA RTX PRO 6000 disponibili in nuove regioni — 16 Jul 2026 xAI lancia Grok Build: un agente di coding interattivo in Rust — 20 Jul 2026 AWS Builder Center: Sandbox gratuiti per workshop — 9 Jul 2026 AWS Lambda introduce lo storage del codice gestito dall'utente su S3 — 16 Jul 2026 Amazon S3: costi ridotti con rimozione vincolo 30 giorni per Standard-IA — 20 Jul 2026 OpenAI lancia una tastiera fisica per la sua app di coding Codex — 16 Jul 2026 Netflix acquista startup AI di Ben Affleck per 587 milioni di dollari — 20 Jul 2026 AWS Network Load Balancer supporta regole listener per routing avanzato — 23 Jul 2026 CloudWatch introduce i Prometheus collectors gestiti — 3 Aug 2026 Cactus Hybrid: Gemma 4 impara a riconoscere i propri errori — 23 Jul 2026 AWS lancia aws-bench: benchmark open source per agenti AI — 27 Jul 2026 Claude Opus 5 disponibile su AWS: un salto di qualità per l'AI — 27 Jul 2026 Amazon Bedrock riduce i prezzi per i modelli OpenAI GPT-5.6 — 3 Aug 2026 Amazon DynamoDB supporta ora la ricerca vettoriale in tempo reale — 6 Aug 2026 Meta lancia Muse Code, un agente AI per grandi basi di codice — 6 Aug 2026 Chip per ripristinare la vista di Science Corporation approvato in UE — 23 Jul 2026 Biblioteche contro la Big Tech: workshop per 'evitare l'AI' — 27 Jul 2026 Un YouTuber famoso ammette un uso eccessivo e dannoso dell'AI — 3 Aug 2026 Top AI researchers lasciano Google per una nuova startup — 6 Aug 2026 Amazon aumenta i prezzi di Echo, Fire TV e Kindle fino al 60% — 24 Aug 2026 AWS Glue 6.0: riduzione prezzi del 30% e supporto Iceberg v3 — 24 Aug 2026 Raccomandazioni di inferenza GenAI in Amazon SageMaker AI Studio — 24 Aug 2026

Updates on its own

Tech Radar

A living excerpt of the Tech Radar: each dot is a pick chosen by a GenAI pipeline and approved by hand — the closer to the center, the more recent. The title takes you to the full page.

27picksFull page

Featured projects

DELFI — GenAI platform for judicial document analysis

Public administration agency · justice sector

A Generative AI platform on AWS for the automatic analysis of large volumes of judicial documents: information extraction and classification with LLM models and semantic search (RAG). I was the Cloud Lead’s operational arm for the technical implementation: infrastructure fully in Infrastructure as Code with CI/CD and end-to-end automated deployment, document-processing services (OCR and language models on GPU instances), monitoring and troubleshooting in production, and a React dashboard for reviewing the results.

  • Generative AI
  • RAG
  • AWS
  • IaC
  • CI/CD

2026

AI & Digital Twin Lab

Higher technical education institute (ITS) · PNRR funds

A teaching lab for Artificial Intelligence and Digital Twin to train students: a prototype on AWS, then dedicated physical infrastructure with GPU nodes. I was the delivery’s operational lead, working largely on my own: systems setup, running the Kubernetes cluster with an MLOps stack (JupyterHub, MLflow), AI demos for supply-chain optimization and anomaly detection, automated backups with a real restore test, and DNSH compliance documentation and acceptance testing.

  • Kubernetes
  • MLOps
  • Digital Twin
  • PNRR

2026

Cyber security & risk management platform

Company in the defense sector

Front-end of an enterprise platform for cyber-security risk analysis and management. I built multi-step wizard interfaces with conditional logic between steps, complex data tables and role-based access control authentication, working in a team on a React stack structured with the Atomic Design pattern. I also contributed to a standardized, reusable front-end codebase for later projects.

  • React
  • TypeScript
  • Redux Toolkit
  • AG Grid
  • Keycloak

2025

matteocarola.com — this website

Personal project · public source code

The site you’re looking at, treated as a production project: Astro 5 with strict TypeScript, Tailwind CSS 4, flash-free light/dark theme, SEO with Open Graph and sitemap, CI on GitHub Actions and zero unnecessary JavaScript. The code is public, with a complete README.

  • Astro
  • TypeScript
  • Tailwind CSS
Source

2026

Behind the scenes

This site isn’t just a portfolio: it’s a small GenAI lab running in production. Two things you can try right now, built from scratch at zero cost.

Ask Matteo — RAG chatbot

An assistant that answers questions about me using only the content of this site: precomputed embeddings, semantic search in a serverless function and an LLM with guardrails. No vector DB, no costs.

  • RAG
  • Embeddings
  • Serverless
  • Gemini
How I built it

Tech Radar — semi-automatic pipeline

A GenAI pipeline that twice a week pulls cloud and AI news from the sources, summarizes it with an LLM and opens a Pull Request. I approve, discard or add a note: nothing publishes itself.

  • GitHub Actions
  • LLM
  • Human in the loop
Go to the radar How I built it

Experience

Software Engineer — Btinkeeng

July 2024 — present

Cloud & AI Engineering

since early 2026

current
  • Implementing and operating AWS infrastructure for Generative AI platforms and AI/ML labs, for public administration and enterprise clients.
  • Infrastructure as Code with modular CloudFormation stacks; CI/CD pipelines on Bitbucket with OIDC authentication, no static credentials, and fully automated deployment.
  • Serverless architectures on AWS (Lambda, Step Functions, API Gateway, DynamoDB, S3, Cognito, CloudFront) and management of EC2 GPU instances for AI workloads.
  • Integration of open-source LLM models and RAG pipelines with semantic search; prompt tuning for document use cases; evaluation of AWS Bedrock from an enterprise perspective.
  • Operations and reliability: monitoring with CloudWatch, root cause analysis with CloudTrail, fixing infrastructure bugs, end-to-end testing on real data.

Enterprise web development

2024 — 2025

  • Development and maintenance of enterprise web applications in React and TypeScript, functional testing and QA, in Agile/Scrum teams (Jira).
  • Projects for clients such as Poste Italiane, EAV and the Umbria Region.

Education & certifications

BSc in Computer Engineering

in progress

Università Mercatorum

2024 — present

AWS Certified AI Practitioner (AIF-C01)

in preparation

Amazon Web Services

2026

Full Stack Web Developer Master

Boolean — 700-hour intensive course

January — July 2024

Technical High School Diploma — Business Information Systems

100/100

Istituto Giancarlo Siani

2017 — 2022

Contact

The quickest way to reach me is by email. I also reply on LinkedIn.

matteocarola.dev@gmail.com
GitHub LinkedIn