Alejandro Rodríguez

1. Data & AI Engineer

   Oct 2025 — Present

   EY (Ernst & Young)

   +90% efficiency in automated document workflows

   Tax automation · Luxembourg client Apr 2026 — Present

   • Backend and service-oriented architecture with Python 3.13 and FastAPI to automate tax workflows and ingest corporate financial statements: Trial Balance, General Ledger and Annual Accounts.
   • Asynchronous workers with MongoDB persistence for run logging, traceability and monitoring; message queues (Redis / Azure Service Bus) and large-scale storage on Azure Blob Storage.
   • Automated generation of final reports ready for regulatory filing, removing repetitive manual work.
   • Contributions to the system frontend with Angular and TypeScript for interacting with the AI components and pipelines.

   Document analysis with Generative AI Oct 2025 — Apr 2026

   • RAG systems (FAISS + LLMs) for automated analysis of 100+ page technical proposals, speeding up decision-making.
   • Automated Word and PPTX report generation engines with Azure OpenAI, improving delivery times by 90%.
   • OCR pipelines (Azure Document Intelligence) and serverless logic (Azure Functions, Docker) for intelligent data extraction.
   • Internal chatbots and advanced prompt engineering (semantic contextualization, per-field weighted boosts in JSON to optimize retrieval with cost-efficient LLMs), >90% test coverage and monitoring in Kibana.

2. AI Researcher (RAG)

   Mar 2025 — Jun 2025

   Grupo NEO · Universidad de Málaga

   −75% inference cost by optimizing LLM context

   • RAG system for generating industrial test cases, served through REST APIs in Flask with SQL persistence and querying (SQLite).
   • Context and resource optimization: −75% spend and −25% LLM response times (Mistral).

3. Technical instructor & academic mentor

   Sep 2023 — Present

   Freelance · UMA

   • Advanced mentoring in algorithm analysis (backtracking, dynamic programming, complexity) and C++ / Python programming.
   • Training in relational design and SQL.

Cybersecurity · Agentic AI

Adyton

MCP server for automated phishing URL triage in SOCs. Seven parallel analysis tools, a Dockerized Playwright agent for deep DOM inspection, and a local LLM that reasons over the grey zone to eliminate false positives.

▸ Case study

Problem

SOC analysts receive hundreds of suspicious URLs per day. Manual triage is slow, and rule-based approaches produce too many false positives in the "grey zone", where a URL is neither clearly legitimate nor clearly malicious.

Solution

I designed an MCP (Model Context Protocol) server exposing 7 typed analysis tools: URL syntax, RDAP/WHOIS, SSL, multi-source reputation, AiTM markers and header auditing. Fast triage runs them in parallel within seconds; deep triage launches a Dockerized Playwright agent (stealth mode) that inspects the live DOM, redirect chains and exfiltration. When the score lands in the ambiguous band, a Decision Engine delegates to a local LLM (Ollama) that reasons like an analyst.

Results

The system turns phishing triage into a composable pipeline interoperable with any MCP client (Claude Desktop, custom agents), cuts false positives through LLM reasoning without sacrificing sensitivity, and produces structured reports (incl. STIX) ready for the analyst.

• Python
• MCP
• Ollama
• Playwright
• Docker

Fintech · Algo trading

ViperTrade

Portfolio management and automated trading platform: Python 3.12 backend (FastAPI), mobile/web app with React Native (Expo) and a quantitative engine built on Pandas and NumPy (RSI, MACD, Fibonacci, Order Blocks) with multithreaded backtesting. DCA automation on Bitget (CCXT) and Pionex with strict risk guardrails, Fernet-encrypted keys in PostgreSQL (Supabase) and a Telegram bot with 1-click order confirmation and contextual analysis via Gemini.

• Python
• FastAPI
• React Native
• Pandas · NumPy
• Supabase
• Gemini

Generative AI · OCR

TicketSaver

Shopping assistant that scans grocery receipts with Google Gemini: extracts merchant, date, total and a categorized product breakdown. Budget tracking, price comparison and shared inventory. PWA and Android app with a Supabase backend.

• React
• TypeScript
• Gemini API
• Supabase
• PWA

Full-stack mobile · Realtime

ApexLap

iOS, Android and web app to track Assetto Corsa lap times, compare live rankings and per-car/track records. Realtime sync and automatic lap upload from the game via a Lua/CSP mod and a Python uploader.

▸ Screenshots

• React Native
• Expo
• Firebase
• TypeScript

More projects