Active Right NowUpdated August 2026

Kyle Barnette

Software Engineer | Backend, Full-Stack, AI & Distributed Systems

I build and operate production software across backend APIs, distributed systems, full-stack applications, and applied machine learning.

Currently working on: Building and operating ChurnLab, a live multi-tenant churn intelligence SaaS built with FastAPI, Next.js, PostgreSQL, Redis, Docker, and explainable machine learning.

Live ChurnLabResumeGitHubLinkedInContact

What I Do

Engineering That Ships and Stays Reliable

Fast scan view of where I spend my time and where I add the most value.

  • Backend APIs and distributed systems
  • Full-stack application development
  • Applied machine learning in production
  • Production operations and reliability

Featured Projects

Production Systems and Engineering Work

Flagship live product first, then research and public engineering work with concrete problem, stack, and delivery tradeoffs.

ChurnLab

Live Product

Live multi-tenant SaaS for churn intelligence, explainable ML, and retention operations

  • Python
  • FastAPI
  • Next.js
  • TypeScript
  • PostgreSQL
  • Redis
  • Docker
  • Linux
  • scikit-learn
  • SHAP
  • Playwright
Problem
Customer-success and RevOps teams need actionable churn signals, not just raw model probabilities.
Solution
Founded, architected, and launched a production SaaS that ingests customer data, trains churn models asynchronously, scores account risk, explains predictions, surfaces revenue exposure, and supports retention workflows.
What I Built
Multi-tenant authentication and authorization, asynchronous Redis-backed scoring, PostgreSQL persistence, explainable ML, risk dashboards, dataset ingestion, retention actions, Stripe integrations, observability, automated backups, CI/testing, and production operations.
Production Engineering
Operate the live application with Docker, Linux, Cloudflare Tunnel/TLS, health checks, structured logging, automated backups, queue recovery, rate limiting, security controls, and end-to-end testing.

Independent Study: TabDiff

Research-backed engineering for synthetic tabular data generation

  • Python
  • NumPy
  • Pandas
  • Jupyter
  • scikit-learn
Problem
Real datasets can be sparse or constrained, making it hard to run safe, repeatable churn experiments at scale.
Solution
Implemented and tested a TabDiff-inspired workflow to generate synthetic tabular records and benchmarked downstream prediction behavior against baseline datasets.
What I Did
Framed the experiment design, implemented preprocessing and evaluation scripts, and documented tradeoffs between fidelity and utility.
Challenges
Controlled for data leakage risk while keeping synthetic outputs statistically meaningful enough for practical experimentation.

3D Rubik's Cube Learning Platform

Full-stack 3D Rubik's Cube simulator and learning platform with React, Three.js, Node.js, Supabase, WebAssembly, and Python solver integration.

  • React
  • Three.js
  • Node.js
  • TypeScript
  • JavaScript
  • Supabase
  • PostgreSQL
  • WebAssembly
  • Python
  • C++
Problem
Cube practice tools are often either toy visualizers or isolated solvers, making it hard to learn, track, and validate solves in one system.
Solution
Built a full-stack learning platform with an interactive 3D simulator, guided tutorials, authenticated solve tracking, leaderboards, and solver-backed API routes.
What I Did
Implemented the React/Three.js simulator and learning flows, Node.js API routes for scramble/solve orchestration, Supabase-backed auth and stats, and solver integration through C++ compiled to WebAssembly plus a Python solver bridge.
Challenges
Kept frontend and backend cube-state logic consistent across cube sizes while coordinating WebAssembly and Python solver paths behind validated API routes.

Skills

Clean Stack Overview

Organized by the tools and domains I use in real project work.

Languages

  • Python
  • Go
  • SQL
  • Java
  • C++
  • C#
  • JavaScript
  • TypeScript

Backend / Systems

  • FastAPI
  • REST APIs
  • gRPC
  • Redis Streams
  • Microservices
  • Distributed Systems
  • Event-Driven Architecture

Frontend

  • React
  • Next.js
  • TypeScript
  • HTML/CSS

Data / ML

  • PostgreSQL
  • Redis
  • Pandas
  • NumPy
  • scikit-learn
  • SHAP
  • explainable ML

Cloud / DevOps

  • Docker
  • Linux
  • Git/GitHub
  • CI/CD
  • AWS
  • Observability

Certifications

  • AWS Certified Cloud Practitioner

Experience

Recent Work

Hands-on engineering across production SaaS, distributed systems, and reliability work.

2025 – Present

Founder & Software Engineer

ChurnLab

  • Founded, architected, launched, and operate a live multi-tenant SaaS for churn intelligence and retention analytics.
  • Built FastAPI/Next.js services backed by PostgreSQL and Redis with asynchronous scoring, explainable ML, authentication, authorization, and production observability.
  • Own production reliability, security hardening, deployment operations, automated backups, performance profiling, and ongoing product development.
Live ChurnLab

2026

Software Engineer Intern

Deepiri

  • Developed backend/distributed-system components using Go, Redis Streams, and gRPC.
  • Worked on event-processing reliability including retries, WAL/DLQ patterns, and latency measurement.

2025 - 2026

Systems Evaluator

Outlier AI

  • Evaluated model and software-system behavior using structured, repeatable testing workflows.
  • Investigated pipeline failures, reproduced edge cases, and documented root causes for faster remediation.
  • Delivered reliability-focused analysis that improved release confidence for downstream stakeholders.

Currently Building

Active Right Now

Current production work, operations, and research in progress.

  • Building and operating ChurnLab as a live production SaaS
  • Performance profiling and observability for asynchronous scoring pipelines
  • Production reliability, security hardening, backups, and deployment operations
  • Continuing applied ML and synthetic-data research alongside the product

Debugging Philosophy

I approach bugs by isolating variables, analyzing logs, reproducing issues, and validating fixes systematically.

Engineering Notes

  • Profiling asynchronous scoring pipelines under production load
  • Hardening multi-tenant authz, backups, and recovery paths
  • Translating model experiments into production-ready service boundaries

Resume

Download a one-page summary of experience, projects, and technical stack.

Download Resume

Contact

Let's Build Something Useful

If you're hiring or building, I can contribute quickly on backend APIs, distributed systems, full-stack delivery, and production operations.

Availability

Actively open to software engineering opportunities across backend, full-stack, platform, distributed systems, and applied AI.

Typical response time: within 24 hours.

Collaboration Fit

  • Open to backend, full-stack, platform, distributed systems, and applied AI roles
  • Comfortable joining existing teams and ramping quickly on production codebases
  • Best fit: projects that value reliability, ownership, and clear communication
EmailLinkedInGitHub