Software Engineer · Fintech & AI
Systems that move money.AI that shows its work.
I build banking-grade Java/Spring systems and LLM applications.
ask my agent: it answers from my real work
01 · The journey
Over 5+ years, one thread: systems people trust.
Six chapters, from banking-grade systems to production AI.
Feb 2020 – Nov 2020
Java Developer · MobyDigital
Started at 18, the same year I began Systems Engineering at UTN. Built monolithic Java/Spring MVC apps and REST APIs, learning Hibernate and MySQL from the ground up.
- Java
- Spring MVC
- Hibernate
- MySQL
Nov 2020 – May 2022
Backend Developer · MobyDigital (client: ICBC / YOY)
Joined the team building ICBC's new digital bank from zero: REST microservices in Java/Spring Boot for digital onboarding (accounts, card issuance), Oracle PL/SQL, and shared libraries used across the platform.
- Java
- Spring Boot
- Microservices
- Oracle PL/SQL
May 2022 – Dec 2022
Full-Stack Developer · MobyDigital
Moved into microfrontend architecture (single-spa and Angular), building independently deployable UI modules and a shared component library with the UX/UI team.
- Angular
- Microfrontends
- single-spa
- TypeScript
May 2022 – Apr 2023
Backend Mentor · MobyDigital
Designed the company's Backend Roadmap and mentored 8+ trainees, including candidates from the government's PIL program. 7 of 8 got hired, several now Senior Devs.
- Mentorship
- Technical Leadership
Jan 2023 – Present
Tech Lead · MobyDigital → ICBC Argentina
Promoted to Tech Lead for YOY's payments and transfers team, later converting to a direct ICBC employee. Maintained 15+ reusable microservices behind transaction flows serving up to 640K logins on peak days. Since Oct 2025, scope expanded into architecture for ICBC's Retail segment: migration roadmaps, C4 diagrams, and ADRs for a platform serving 560K+ monthly active users.
- Leadership
- Payments
- Microservices
- Architecture
- ADRs
2025 – Present
Independent project · Gastra
Building a restaurant-recommendation app on the side: hybrid retrieval (pgvector + geolocation + weather) that turns a natural-language query into a ranked result via LLM, hexagonal architecture, local inference via Ollama.
- RAG
- LLM
- Spring AI
- pgvector
- Hexagonal Architecture
02 · Selected work
Case studies, not screenshots.
I've worked on a variety of projects, from banking-grade systems to production AI.
Product · AI
Gastra
A restaurant-recommendation app with hybrid retrieval (vector search, geolocation, live weather) that turns a natural-language query into a ranked list or a single high-confidence match. Hexagonal architecture, local LLM inference, built under Spec-Driven Development.
Platform · Banking
Banking at scale
Tech Lead for YOY, ICBC's mobile banking app: transaction flows serving up to 640K logins on peak days. Core microservices at scale: fraud prevention (~35M monthly hits), transfers (~13M), bill payments (~4.5M).
Engineering · This site
degiorgi.dev
This site: a static Next.js shell with a RAG-style console over my own work. Every answer traces back to the source chunk it came from.
03 · Stack
// languages
- Java
- Python
- TypeScript
// frameworks
- Spring
- Next.js
- LangChain
- LangGraph
- scikit-learn
// ai tools & models
- Claude Code
- Gemini
- Cursor
- v0
- Ollama
// data
- PostgreSQL
- MySQL
- Oracle SQL
// infra & devops
- Docker
- Kubernetes
- OpenShift
- GitLab
- GitHub
- Apache Maven
- Bruno
// observability
- Elastic
- Kibana
- Dynatrace
// tools change. The discipline doesn't.
Let's talk.
I like talking shop: fintech, distributed systems, AI engineering. If that's you too, reach out.
GET /contact →200 OK