Overview
GitHub Contributions
I’m Mateo, an Electrical & Computer Engineering student at Cornell building the infrastructure that serves large models—and the research that stress-tests them.
01InfrastructureView ↗02ResearchView ↗03BuilderView ↗
PrivateMind
OpenAI-compatible batch inference, GPU load management, and production text-to-image on B200/B300s.
Open Weight, Open Risk
A training-free jailbreak that steers a target model’s chain-of-thought to expose shallow alignment.
Dispatch
Agentic pentesting platform—and Best Developer Tool at the 2026 Cornell AI Hackathon.
Blog(1)
- closed-weight ECIopen-weight ECIBCSI (this paper)forecast
Capability curves use Epoch AI ECI for major closed- and open-weight releases, plotted at official release dates. The BCSI series is this post's Bio/Chem Safety Index (BCSI = 1 − mean(BRI, CRI)) under the DeepSeek judge; the hatched region marks BCSI below 0.5.
Stack
Infrastructure
Metrics
Experience
Options Technology
- Location
- New York, New York
- Location type
- (On-site)
Software Engineering Intern, PrivateMind Platform
- Employment Type
- Internship
- Employment Period
- 06.2026—08.2026
- Duration
- 3m
Software Engineering Intern, PrivateMind Platform
- Owned and shipped an OpenAI-compatible Batch API end-to-end (async worker draining against GPU capacity, per-request billing, DB schema, SDK, Helm deployment), backed by ~7,000 lines of contract tests; integrated Pure Storage FlashBlade S3 results storage for large inference artifacts.
- Benchmarked frontier LLMs before production serving: Poisson-distributed concurrency sweeps measuring time-to-first-token, inter-token latency, end-to-end latency, pre-emption, and throughput at p50/p95/p99 across multi-GPU HGX deployments.
- Owned the gateway load-management layer (priority scheduling, per-model streaming timeouts/keepalive, per-key concurrency limits, token-budget admission control gating batch work on GPU memory pressure) on NVIDIA B200/B300 GPUs.
- Brought a novel text-to-image model through full production deployment (GPU configuration, gateway routing, bounded concurrency, per-image pricing, UI) and built the platform content-safety layer.
- Merged 133 PRs across 10 repos (Rust, TypeScript, Python, SQL, Helm); unified two divergent agent execution paths, deleting ~6,000 lines of legacy code.
- Linux
- Kubernetes
- OpenShift
- Helm
- vLLM
- SGLang
- Pure Storage FlashBlade S3
- Rust
- TypeScript
- Python
- PostgreSQL
Cornell University
- Location
- Ithaca, New York
- Location type
- (On-site)
Engineering Manager & ML Engineer, Generative AI at Cornell
- Employment Type
- Part-time
- Employment Period
- 10.2025—05.2026
- Duration
- 8m
Engineering Manager & ML Engineer, Generative AI at Cornell
- Led an 8-person engineering team building an ESG risk monitoring platform for Investcorp ($60B+ AUM).
- Implemented a vendor due-diligence agentic pipeline with fintech company QuickFi.
- Collected vendor data from government websites, regulatory filings, and public databases; reconciled it into consolidated profiles and generated written due-diligence reports.
- TypeScript
- Python
- Agentic Pipelines
- Engineering Management
AI Alignment Undergraduate Researcher, Cornell NLP Group
- Employment Type
- Part-time
- Employment Period
- 10.2025—05.2026
- Duration
- 8m
AI Alignment Undergraduate Researcher, Cornell NLP Group
- Built the evaluation pipeline behind "Open Weight, Open Risk" (co-author): 15 frontier open-weight models across 800 WMDP-derived adversarial bio/chem requests, run locally on B200-class GPUs with vLLM and SGLang.
- Co-designed the training-free "inception" attack, in which a small uncensored 7B architect model pivots target reasoning and lifts compliance from a ~10% baseline to >95% under 5 iterations, exposing "shallow alignment" in open-weight reasoning models, with a PhD researcher at the Cornell NLP Group.
- Python
- vLLM
- SGLang
- LLM Safety
- Evaluation
Hospital for Special Surgery
- Location
- New York, New York
- Location type
- (On-site)
Machine Learning Research Intern
- Employment Type
- Internship
- Employment Period
- 06.2024—08.2024
- Duration
- 3m
Machine Learning Research Intern
- Analyzed immunological datasets with Dr. Amit Lakhanpal and generated protein embeddings using Meta's ESM model to support rheumatoid arthritis treatment research.
- Python
- ESM
- Protein Embeddings
- R
Education
Cornell University, College of Engineering
- Employment period
- 08.2024—12.2028
- Degree
- B.S.
- Field of study
- Electrical & Computer Engineering
Cornell University, College of Engineering
- Bachelor of Science in Electrical & Computer Engineering, GPA 3.75 / 4.00. Expected graduation: December 2028.
- Coursework: CS 4414 (Systems Programming), ECE 3140 (Embedded Systems), ECE 2300 (Digital Logic and Computer Organization), MATH 2940, PHYS 2213, MATH 2930.
- C
- C++
- CUDA C++
- Python
- Rust
- Computer Architecture
- Machine Learning
Projects(4)
Dispatch
- Period
- 03.2026—Present
Agentic penetration testing platform that turns vulnerability findings into ready-to-merge GitHub PRs. / Best Developer Tool — Cornell AI Hackathon 2026
- Orchestrated security agents using Mastra and OpenRouter, with isolated execution via Blaxel Sandboxes
- Designed Slackbot and Datadog middleware for triggering scans and ingesting logs to surface findings
- TypeScript
- Mastra
- OpenRouter
- Blaxel Sandboxes
- Datadog
- Slack API
- Agentic Systems
- Security
Open Weight, Open Risk
- Period
- 10.2025—Present
Training-free jailbreak in which an uncensored model steers the target's chain-of-thought. / Publication
- Achieves near-universal compliance across open-weight models (DeepSeek, Kimi, GLM, Qwen)
- Iterative prompt injection and LLM-as-judge pipelines to streamline evaluation
- Evaluated 15 frontier models across 800 WMDP-derived biosecurity and chemical security requests
- Python
- vLLM
- SGLang
- LLM Safety
- Red Teaming
- Evaluation
- Research
Debately
- Period
- 05.2025—Present
Moderated debate platform with fact-checking pipelines and an LLM-driven consensus tracker (Anthropic API, Supabase, Vercel/Railway). / Source
- Engineered fact-checking pipelines with web data normalization, preventing hallucinations in consensus
- TypeScript
- React
- Anthropic API
- Supabase
- Vercel
- Railway
TensorAtelier
- Period
- 01.2025—Present
Pluggable PyTorch interface with automatic optimization/profiling and configurable multi-accelerator support, reducing boilerplate in ML research. / Source
- Python
- PyTorch
- CUDA
- Profiling
- ML Infrastructure
Awards(3)
Cornell AI Hackathon 2026
- Prize
- Best Developer Tool
- Awarded in
- Received in Grade
- University
- Built Dispatch, an agentic penetration testing platform automating vulnerability detection and remediation (Pitch Deck · GitHub)
- Developed integrations with Datadog, Slack, and Linear for seamless integration with developer workflows
Regeneron WESEF 2025
- Prize
- 2nd Place — Computer Science
- Awarded in
- Received in Grade
- High School
Regeneron WESEF 2024
- Prize
- 2nd Place — Computer Science
- Awarded in
- Received in Grade
- High School
Bookmarks(2)
The Techno-Optimist Manifesto
- Author
- Marc Andreessen
- Category
- Article
- Bookmarked on
The Bitter Lesson
- Author
- Rich Sutton
- Category
- Article
- Bookmarked on