Available May 2027 · Boston / NYC / Houston

Nate Cirino.
Software Engineer.

I work across the stack: distributed systems in Rust, applied AI in production, and the infrastructure to ship it all at scale.

Explore the workGitHubLinkedIn
p95 latency
0.0ms
throughput
0k ops/s
uptime
0.00%
commit log
0idx
01 / work

Selected work

Each panel below is live: running in your browser, no backend. Poke at them.

Kepler

In progress

A distributed, linearizable key-value store in Rust. From-scratch LSM-tree storage engine, from-scratch Raft consensus, MVCC transactions.

Mar 2026 - present
wal0
memtable · sorted0/8
flushed
L0flushed, may overlap
empty
L110× larger
empty
L2100× larger
empty
term
1
leader
n0
state
ok
replicated log
awaiting writes…

Real kepler-raft compiled to WebAssembly, running in your browser, not a simulation. Write a key to append to the leader's log; kill the leader to watch a genuine re-election.

  • Built the storage engine from the ground up: memtable, write-ahead log, and leveled SSTable compaction.
  • Implemented Raft consensus from scratch: leader election, log replication, and membership changes.
  • MVCC transaction layer providing snapshot isolation and linearizable reads across the cluster.
RustRaftLSM-treeMVCCTokio

Sentinel

Full-stack · AI

An AI SRE that triages incidents. It autonomously investigates real telemetry (metrics, traces, logs, recent deploys, runbooks), then streams a root-cause hypothesis with cited evidence and a human-approved remediation.

Apr 2026 - present
◎incident #1042sev1Checkout failing with 504s
agent · streamingalert blames: gateway
awaiting investigation
eval harnessroot-cause accuracy across labeled incidents
5/5 · 100%avg 4.4 toolsavg 5.0s
Replay of real investigations · live system runs on Docker: Claude tool-use over Prometheus / Tempo / Loki + pgvector RAG. Eval numbers are a representative run of the in-repo harness.
  • Streaming agentic loop with tool-use over Prometheus / Tempo / Loki: real PromQL, TraceQL, and LogQL, not mocks.
  • RAG over runbooks and postmortems in pgvector, with prompt caching on the frozen system prompt + tool schemas.
  • Human-in-the-loop remediation gated behind approval, plus an eval harness scoring root-cause accuracy across labeled incidents.
  • Running in SculptAI's production infrastructure, where it actively triages incidents today.
Next.jsFastAPIClaudepgvectorOpenTelemetryDocker

NOAA AUV Sonar Detection

ML · Inference

Sonar anomaly-detection models for NOAA autonomous underwater vehicles, optimized for custom-silicon real-time inference.

Apr 2025 - Jul 2025
hydrophone spectrogram · 0–24 kHz · live anomaly detection
  • Trained anomaly-detection models over raw hydrophone sensor data.
  • Optimized the inference pipeline with the AWS Neuron SDK for custom-silicon deployment, cutting real-time latency.
  • Designed the feature-engineering pipeline alongside NOAA researchers.
PythonPyTorchAWS Neuron SDKSageMaker
02 / experience

Where I've shipped

Wayfair

Software Engineering Co-op● incoming
Jul 2026 - Dec 2026 · Boston, MA
  • Incoming co-op on the Supply Chain team, building logistics systems that move products from suppliers to warehouses to customers across international transportation and delivery networks.

Wolters Kluwer

Software Engineer● current
Jul 2025 - Jun 2026 · Boston, MA
  • Integrated UpToDate Expert AI into CME accrual/redemption microservices, expanding GenAI clinical decision support to 100K+ clinicians.
  • Built a Java Spring microservice for CME credit issuance inside Expert AI's clinician-facing workflow.
  • Overhauled Relay JMS broker serialization with json-io across 20+ microservices, eliminating timeouts at scale.
  • Strengthened Expert AI search and filtering logic, improving clinical query performance and relevance.
  • Implemented OIDC/IdP authentication via Azure for new and existing consumers, securing identity access.

SculptAI

Software Architect● current
Jan 2025 - present · New York, NY
  • Led an engineering-org restructure at a $1M-funded startup, partitioning teams into DevOps, product dev, and agentic-workflow squads.
  • Architected a Flask backend with a RAG pipeline for personalized workout recommendations on Amazon Bedrock Knowledge Bases.
  • Evolved the RAG stack from LangChain → Vespa → zero-ops Bedrock KB, improving retrieval quality while cutting operational overhead.
  • Maintained 99.99% uptime for 2K+ users via AWS WAF, Datadog, and GitLab / Kubernetes / Terraform.

Chevron New Energies

Software Engineer Intern
May 2024 - Aug 2024 · Houston, TX
  • Built RESTful APIs and data pipelines with Azure Data Factory, improving product efficiency 40%.
  • Implemented CI/CD with Azure DevOps and Kubernetes, enabling automated deployment and rollback.
  • Integrated PyTorch models with the AWS Neuron SDK for ML inference supporting 3M+ consumers globally.

Legacy Community Health

Software Engineer Intern
May 2023 - Aug 2023 · Houston, TX
  • Debugged and resolved defects in production Epic (EHR) modules, helping sustain 99.9% application uptime.
  • Reviewed code and enforced engineering standards across healthcare applications serving 100K+ patients.
03 / stack

Tools & Frameworks

Languages
RustTypeScriptPythonCC++JavaJavaScriptBash
Infrastructure
AWSAzureDockerKubernetesTerraformJenkinsGroovyLinux
Frameworks & Data
ReactNext.jsNode.jsExpressPyTorchTensorFlowPostgreSQLPandas
Northeastern University
B.S. Computer Science, Systems Concentration
GPA 3.6 · Dean's List (all semesters) · Boston, MA · May 2027
json-io OSS ContributorTech Lead, Northeastern Electronic RacingPawHacks
04 / contact

Building something hard? Let's talk.

Fastest way to reach me is email.

cirino.na@northeastern.eduGitHubLinkedIn