Skip to content
View ffumero2003's full-sized avatar
1 follower · 0 following

Achievements

Achievement: Pull SharkAchievement: Pair ExtraordinaireAchievement: YOLOAchievement: Quickdraw

Achievements

Achievement: Pull SharkAchievement: Pair ExtraordinaireAchievement: YOLOAchievement: Quickdraw

Block or report ffumero2003

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ffumero2003/README.md

Felipe Fumero

Data & AI Engineer — batch and streaming pipelines (BigQuery, dbt, Airflow, PySpark) and production LLM applications, including work on the Claude API.

Currently: building out end-to-end data platforms with governance and IaC baked in, not bolted on.

Stack

Ingestion & Processing Warehouse & Modeling Orchestration & Infra AI / LLM
Python, PySpark, pandas BigQuery, dbt, SQL Airflow, Docker, Terraform Claude API, RAG, MCP

Featured projects

  • github-firehose-activity — PySpark pipeline processing 100M+ GitHub Archive events into a BigQuery star schema, dbt-tested, Airflow-orchestrated
  • terraform-gcp-data-platform — A GCP analytics platform where every resource is IaC, and an AI viewer on top is infrastructurally incapable of writing data
  • credit-risk-explainable-ai — Credit-risk scoring (XGBoost + SHAP) with Claude-generated, FCRA-style explanations and fair-lending guardrails
  • retrieval-benchmark — Measured retrieval quality: naive vector search vs. hybrid BM25/vector + rerank, over a real 10-K filing
  • b2-mcp-server — MCP server exposing Backblaze B2 as tools an AI client can call, with a fenced filesystem and audit-logged deletes
  • b2-sync — Backup CLI in Go with SHA-256 diffing and a bounded worker pool

📫 LinkedIn · felipefumero.com

Pinned Loading

  1. credit-risk-explainable-ai credit-risk-explainable-ai Public

    This project ingests the Home Credit Default Risk dataset into Google BigQuery, transforms it into clean analytical models with dbt, trains an XGBoost credit-risk model (benchmarked against a logis…

    Python 1

  2. github-firehose-activity github-firehose-activity Public

    PySpark pipeline processing 100M+ nested JSON events from the GitHub Archive firehose — explicit schema, multi-grain star schema (event + commit facts) in BigQuery, dbt tests, and Airflow orchestra…

    Python 1

  3. retrieval-benchmark retrieval-benchmark Public

    Measured retrieval-quality benchmark: naive vector search vs contextual chunking + hybrid BM25/vector + Cohere rerank, over Apple's 10-K. FastAPI and React side-by-side diff, with precision/recall/…

    Python 1

  4. b2-mcp-server b2-mcp-server Public

    MCP server exposing Backblaze B2 Cloud Storage as tools an AI client can call — list, upload, download, delete, and report bucket usage against a budget. TypeScript, stdio, filesystem access fenced…

    TypeScript 1

  5. b2-sync b2-sync Public

    Backup CLI in Go: watches a folder, diffs against a JSON state file by mod-time and SHA-256, and uploads only what changed to Backblaze B2 through a bounded worker pool with retry. Deletion is opt-in.

    Go 1

  6. terraform-gcp-data-platform terraform-gcp-data-platform Public

    A minimal analytics platform on Google Cloud where every resource is provisioned as code with Terraform, and a Streamlit + AI viewer on top can read the data but physically cannot write it — becaus…

    Python 1