CS student building cloud systems and measuring them honestly. Looking for a SWE internship: cloud, infrastructure, security.
I build systems end to end, deploy them on AWS, test them, and write down what the numbers actually say. Every number on this page comes from a test run, a live measurement or a merged PR you can open. If a project did not work, its README says so in the first paragraph.
Python daily. AWS SAM, Step Functions, Lambda, DynamoDB, IAM. C when speed matters. CI on every repo.
Agentic NetOps: closed-loop network remediation on AWS
A serverless platform that heals a VPC when it drifts from declared intent: detect, prove, measure, generate a repair, approve, execute, verify. The LLM proposes and deterministic systems dispose, so control flow runs on config diffs and never on model output. Repairs are bounded to converge-to-baseline by a versioned policy gate, routed through fail-closed autonomy tiers (auto-execute behind a veto window, one approval, or two-party maker-checker), and executed under per-incident scoped STS credentials. Every decision lands in a hash-chained ledger anchored out of band.
- Measured on the live deployment: self-healed a seeded fault in 191 s with no human touch, and all 9 seeded fault classes remediate.
- 301 tests in credential-free CI, 18 architecture decision records, a security findings register from repeated adversarial audits, $0 idle.
- The code is private. I'm happy to demo it or walk through the code, just email me.
AWS SAM Step Functions Lambda DynamoDB EventBridge Cognito Bedrock IAM CloudFront
walkforward: volatility, sizing and purged walk-forward splits in C11, with a Python front door
pip install walkforward
GARCH(1,1), EWMA and range-based volatility, Kelly and vol-target sizing, purged walk-forward splits and ridge regression, written from scratch in C11 with a ctypes binding over numpy and pandas. A 12-job CI matrix runs gcc, clang, macOS, Windows, 32-bit, sanitizers and a fuzz sweep, then checks the C against pandas, numpy, scikit-learn and arch. Wheels for five platforms ship through cibuildwheel and PyPI trusted publishing, with no token anywhere. The worked SPY example reports honest out-of-sample numbers, including the ones that did not help.
C11 CMake ctypes cibuildwheel GitHub Actions PyPI trusted publishing
AsOf: a market intelligence engine that grades itself
Collects signals from SEC Form 4 insider buys, 8-K events, your own monitors and any local tool, seals each one at the instant it became public, then grades it against SPY at 5, 20 and 60 trading days. Entry is the close of the first session that ended after the filing's acceptance time, DST-aware. Imported history can never reach the scorecard, small samples are withheld rather than shown, and grades are written once and never revised. Runs locally on free, keyless data. 103 offline checks in CI.
TypeScript Next.js SQLite SEC EDGAR
Retro: offline voice remote for Spotify
Speech never leaves the machine: Vosk and faster-whisper on the GPU, your own artists and playlists injected into the recognizer vocabulary, OAuth PKCE with tokens stored on device, and a Windows tray app with a one-command installer.
Python Vosk faster-whisper OAuth PKCE
Three quant systems, each built to find out whether an edge existed, each shut down by its own pre-committed gate. The READMEs lead with the negative result.
- density-desk: an autonomous Kalshi trading desk that tested three hypotheses against live markets with real money. All three failed, for about $6 in realized losses. 112 tests.
- technical-engine: a 506-ticker intraday scanner with Discord alerts and a look-ahead-safe backtest. 5,400 alerts over 60 sessions came to -748R, and three rescue hypotheses were tested and rejected. 99 tests.
- alphaos: a $0 point-in-time alpha research lab with an append-only trial ledger and Deflated Sharpe. 8 campaigns, 34 pre-registered hypotheses, 0 survivors. 107 tests.
Merged upstream:
| Repo | PR | Fix |
|---|---|---|
| moto | #10234 | SQS raises PurgeQueueInProgress on a repeated purge within 60 s |
| statsmodels | #10229 | Probit log-likelihood, score and Hessian computed in log space |
| nautilus_trader | #4914 | AroonOscillator lowest-low scan and tie handling |
| pandas_market_calendars | #485 | TASE calendar updated for the Monday to Friday trading week |
| anndata | #2652 | Raise on float index arrays with any non-integer value |
| scikit-bio | #2580 | to_indices and to_cigar on single-segment alignment paths |
| PyLabRobot | #1273 | Serialize Container.no_go_zones so copy() and deserialize() keep them |
In review: mypy #21926 (report irrefutable match patterns that make later cases unreachable) and biotite #946 (CONECT records for bonds across insertion codes).
Four public Roblox games, built with Rojo, Studio automation and git. Roblox profile.
- B.S. Computer Science, California State University San Marcos, 2026 to present
- AWS Certified Cloud Practitioner
- Collaborator and apprentice, Riffyx Labs mentorship (Jun 2025 to Jun 2026): AWS fundamentals, debugging and Git workflows alongside Cisco professionals