Switch to English?
Yes
Переключитись на українську?
Так
Переключиться на русскую?
Да
Przełączyć się na polską?
Tak

Ignazio De Santis

AI systems that work in production — not just demos.
Offer Ignazio work on your next project.

Italy Padova, Italy
2 months 6 days back
Available for hire available for hire
age 29 years
on the service 2 months 7 days

Rating

Successful projects
No data
Average rating
No data
Rating
278
Python 1
928 place out of 4464
AI & Machine Learning 1
337 place out of 2836

CV

Software Engineer specializing in AI systems, Python, and production-grade automation. I build things that hold up under real conditions — not just demos.

What I build:

  • RAG pipelines & LLM applications — retrieval-augmented generation systems with multi-cloud AI backends (Vertex AI, OpenAI), vector database integration, and FastAPI REST APIs
  • AI agent frameworks — durable execution layers with human-in-the-loop approvals, retries, and audit trails
  • Developer tooling — evaluation harnesses for prompts/agents, code scaffolding generators, debugging assistants
  • Data & analytics — pipeline monitoring, schema drift detection, marketing attribution models
  • Full-stack applications — Python backends, Next.js frontends, Rust systems components

Tech stack: Python · FastAPI · LLMs (OpenAI, Gemini) · RAG · LangChain · Vector DBs · Next.js · TypeScript · Rust · PostgreSQL · Docker

Available for remote contract work. Fast turnaround, clean code, production-ready.

Skills and abilities

Portfolio


  • Revenue Signal Copilot — AI Lead Scoring System

    AI & Machine Learning
    A lead intelligence and prioritization system that scores accounts using public signals and internal notes. Helps sales teams focus on the highest-probability deals by surfacing buying intent signals and account health indicators. Built with Python and AI.

    #python #sales #automation #ai #businessintelligence #leadscoring #crm #saas
  • Data Quality Watchtower — Pipeline Monitoring

    Python
    A monitoring assistant that detects schema drift, anomalies, and suspicious dataset changes before pipelines break. Catches bad data early so engineers don't discover problems in production. Supports validation rules, alerts, and anomaly detection. Built with Python.

    #python #monitoring #dataengineering #datavalidation #anomalydetection #dataquality #observability
  • EvalOps Workbench — LLM Prompt & Agent Evaluation

    AI & Machine Learning
    A local-first evaluation harness for prompts, tools, and AI agents with regression tracking and experiment history. Run structured evals, compare prompt variants, and catch regressions before shipping. Essential for teams building serious LLM-powered products. Built with Python.

    #python #llm #evaluation #promptengineering #aiagents #developertools #testing #mlops
  • Spec-to-Agent Scaffold — AI Code Generator

    AI & Machine Learning
    A generator that turns structured product specs into typed Python agent service skeletons. Eliminates boilerplate for AI agent projects — just describe what your agent should do, and get a production-ready scaffold with correct types, structure, and wiring. Built with Python.

    #python #ai #codegen #scaffolding #agents #developertools #automation
  • Customer Ops Briefing Bot — AI Account Intelligence

    AI & Machine Learning
    A weekly intelligence agent that turns CRM activity, meeting notes, and support history into concise account briefs. Helps customer success and sales teams stay on top of their accounts without manual work. Built with Python and AI agents.

    #python #automation #crm #ai #aiagents #businessintelligence #customersuccess #sales
  • Long-Form Content Intelligence Engine

    AI & Machine Learning
    A system that ingests PDFs, web pages, and notes, then answers questions with citations and a verification step that flags uncertain claims. Ideal for research, legal, and compliance use cases where source grounding matters. Built with Python and LLMs.

    #python #ai #llm #nlp #rag #research #contentintelligence #citations
  • Marketing Attribution — Multi-Touch Analytics

    Python
    Multi-touch marketing attribution analysis comparing last-click, linear, and time-decay models across 17,000+ customer touchpoints. Helps marketing teams understand which channels actually drive conversions. Built with Python, Pandas, and Jupyter Notebook — ready for integration into existing analytics pipelines.

    #python #pandas #dataanalysis #marketing #attribution #analytics #datascience #jupyter
  • Repo RAG Debugger — AI-Powered Debugging Assistant

    AI & Machine Learning
    A source-aware debugging assistant that indexes codebases, stack traces, and documentation to propose grounded fixes. Uses RAG to retrieve the most relevant code context before generating a diagnosis. Reduces debugging time by grounding answers in the actual repository. Built with Python, FastAPI, and embeddings.

    #python #fastapi #rag #debugging #ai #embeddings #developertools #llm
  • Agent Runbook Orchestrator — AI Operational Automation

    AI & Machine Learning
    A durable execution layer for AI-assisted operational runbooks with human-in-the-loop approvals, automatic retries, and a full audit history. Enables teams to define repeatable ops procedures that AI agents can execute step-by-step, pausing for human approval on sensitive actions. Built with Python and FastAPI.

    #python #fastapi #automation #agents #aiagents #runbooks #humanintheloop #durableexecution
  • SentinelID — Passkey Face Login System

    AI & Machine Learning
    A security-focused authentication system that replaces passwords with a passkey-style face login. Built as a desktop app with a companion web dashboard. Performs on-device computer vision and biometric recognition — no data sent to the cloud. Built with Python, Rust, Next.js, and face recognition libraries.

    #python #rust #nextjs #authentication #biometrics #facerecognition #desktopapp #security
  • NexusRAG — Multi-Cloud RAG Platform

    AI & Machine Learning
    A robust, scalable RAG (Retrieval-Augmented Generation) service with a pluggable architecture supporting multiple cloud AI backends (Google Vertex AI and others). Features multi-modal output (text + speech), vector database integration, and a FastAPI-based REST API. Built with Python, FastAPI, and cloud-native AI services.

    #python #fastapi #rag #ai #llm #vertexai #vectordatabase #machinelearning