Skip to content
View santinomarial's full-sized avatar
πŸ’­
Building
πŸ’­
Building

Block or report santinomarial

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.

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
santinomarial/README.md

Hi, I'm Santino Marial πŸ‘‹

Software Engineer Β· CS & Statistics @ Harvard ('27)

I build backend systems, ML pipelines, and AI infrastructure β€” from a C++ database engine written from scratch to AWS-based cybersecurity platforms processing 10,000+ alerts a day. I care about performance, reliability, and shipping things people actually use.

πŸ“ Cambridge, MA πŸ“« smarial@college.harvard.edu

LinkedIn GitHub Email


Tech Stack

Languages C++ Python Java TypeScript JavaScript SQL

Backend & Data FastAPI Flask Node.js GraphQL PostgreSQL MongoDB Redis

ML & Infra NumPy Pandas scikit-learn React Docker AWS Git


Featured Projects

Project Description Tech
High-Performance Database Engine Database engine built from scratch with buffer pool management, B+-tree indexing, write-ahead ACID logging, and concurrent transactions β€” 4x query speedup, 35% less memory vs. baseline. C++
Collaborative Code Editor Real-time editor for 30+ concurrent users with sub-150ms sync via a custom Operational Transformation engine; horizontally-scalable WebSockets with Redis pub/sub. Adopted by 25 Harvard CS students. Node.js Β· React Β· WebSockets Β· Redis Β· MongoDB
Forward Repair for RAG Pipelines DSPy-based RAG repair framework that corrupts and repairs specific LM pipeline stages, comparing query-stage vs. answer-stage failures with HotpotQA + BM25 retrieval. Python Β· DSPy

πŸ’‘ Pin these repos on your profile (Profile β†’ Customize your pins) so they render as cards too.


Experience

  • Machine Learning Fellow β€” Harvard Undergraduate Machine Intelligence Community (Feb 2026 – Present)
  • Applied AI Fellow β€” Harvard Grid (Sep 2025 – Dec 2025)
  • Software Engineer Intern β€” MoMo from MTN (May 2025 – Aug 2025)
  • Engineering & Research Intern β€” diiVe, Cape Town (Jun 2024 – Aug 2024)
  • Program Coordinator & Engineer β€” Young Leaders Academy of South Sudan (Sep 2023 – Present)

Pinned Loading

  1. PrepOS PrepOS Public

    Adaptive technical interview prep tracker with spaced repetition engine, weak-pattern detection, and personalized recommendation algorithm. Built with FastAPI + PostgreSQL + React.

    Python

  2. random-mst-scaling random-mst-scaling Public

    Empirical validation of MST weight scaling across random graph models using Kruskal's algorithm.

    Python

  3. Meridian Meridian Public

    TypeScript 1

  4. forward-repair-lm-pipelines forward-repair-lm-pipelines Public

    Python