</> raynercodes </>
Portrait of Leo Rayner

Baxley, Georgia

Leonardo Rayner

RaynerCodes

AWS Certified Developer Associate | Backend & Cloud Engineer | Founder, LeverageCut and RaynerCodes Cloud Solutions | Python • SQL

  • Python
  • FastAPI
  • AWS Lambda
  • DynamoDB
  • API Gateway
  • Docker
  • Stripe
See my Experience/Projects Resume/Certifications Get in touch with RaynerCodes

Resume

Summary

Independent Backend Engineer, founder of RaynerCodes Cloud Solutions, and creator of LeverageCut — a B2B SaaS in active beta. Builder of production systems that have measurably cut Lambda request duration by 53%, reduced secrets-management costs by ~100%, and increased extraction pipeline success rates from 27.6% to 72.4%. Specializing in Python and serverless AWS architecture with hands-on depth in security-first API design, event-driven systems, and Infrastructure as Code. AWS Certified Developer Associate and Cloud Practitioner.

Certifications

Experience

RaynerCodes Cloud Solutions — Independent Backend Engineering & Consulting

Jul 2026 — Present

Backend Engineer (Contract) · Debo's Boxing and Fitness — Client Project

  • Architected and deployed a full-stack serverless RESTful API and booking platform using Python, FastAPI, AWS Lambda, DynamoDB, and API Gateway — handling real-time scheduling, payment processing, and business logic for a real paying client
  • Built end-to-end Stripe payment integration including webhook-driven order confirmation, idempotent payment handling, and automated conditional refunds based on real business rules
  • Resolved a distributed concurrency bug where concurrent booking requests caused double-bookings — implemented an atomic conditional-write claim mechanism in DynamoDB, eliminating the race condition without locks or external coordination
  • Optimized an AWS Lambda authentication workload, reducing request duration by 53% (2,334ms → 1,094ms) by tuning memory allocation (256MB → 1024MB) for a security-critical PBKDF2 password hashing operation at 600,000 iterations — with negligible cost impact due to Lambda's memory×duration billing model
  • Identified and eliminated an over-provisioned per-secret KMS key architecture (15 dedicated keys across three environments), migrating to AWS Systems Manager Parameter Store and cutting monthly secrets-management costs by ~100% (~$21 → ~$0) with zero downtime across live production environments

Founder & Backend Engineer · LeverageCut — B2B SaaS competitor price-monitoring platform, active beta

  • Redesigned a synchronous data pipeline to an async event-driven SQS worker architecture with partial-batch-failure handling and dead-letter queue retry policy — eliminating 100% of API Gateway timeout failures since deployment
  • Integrated Google Gemini API as an AI-powered extraction layer alongside a rules-based scraping pipeline, using structured JSON schema validation, confidence scoring, and retry logic distinguishing transient from permanent failures
  • Built a defense-in-depth web scraping pipeline (JSON-LD, microdata, platform-specific parsing, and Bright Data headless rendering fallback) — measurably increasing successful price extraction from 27.6% to 72.4% across a 29-site real-world test of JavaScript-heavy, anti-bot-protected e-commerce sites
  • Diagnosed and resolved an intermittent production Lambda error by tracing an unclosed async HTTP client's garbage-collection timing against asyncio event loop teardown, resolving it via the SDK's own context-manager cleanup pattern

Projects

Fintech Serverless API (Loan Lending Platform)

Feb 2026 — Present

Live Demo · GitHub Repository

  • Designed and built a serverless backend system on AWS Lambda, API Gateway, DynamoDB, and SQS FIFO implementing credit score based auto decisioning that auto-rejects high-risk applicants without human intervention — eliminating manual review overhead while enforcing one-way loan lifecycle state transitions that prevent invalid status reversals
  • Implemented a security-first architecture with blast radius minimization using 6 separate customer managed KMS keys, AES-256-GCM application-level encryption, and PBKDF2-HMAC-SHA256 password hashing with 600,000 iterations and a dedicated pepper key
  • Built an immutable compliance audit trail using S3 Object Lock in Compliance mode with 3-year retention, logging 100% of lifecycle events with tamper-proof retention and JWT-extracted actor identity — ensuring every loan decision is attributable and immutable for regulatory compliance
  • Implemented a CI/CD pipeline with GitHub Actions, OIDC federation, CodePipeline, and CodeDeploy canary deployments with automated CloudWatch rollback — achieving zero unplanned production rollbacks by catching regressions in staging before reaching real users
  • Wrote 80+ automated unit and integration tests using Pytest and Moto covering credit decisioning boundaries, lifecycle state transitions, idempotency, and authentication flows — load tested with Locust (8 concurrent users, 2-minute run) achieving 99.3% success rate (405/408 requests) with zero throttling

Open Source

SlowAPI — Python rate limiting library for FastAPI and Starlette. PR #279: resolved a Starlette 1.0 incompatibility by replacing the deprecated @app.route() with app.add_route() across the test suite and documentation.

Skills

Languages

  • Python
  • SQL

Cloud / AWS

  • Lambda
  • API Gateway
  • DynamoDB
  • SQS
  • S3
  • KMS
  • Secrets Manager
  • IAM
  • CloudFormation
  • SAM
  • EventBridge
  • CodePipeline
  • CodeDeploy
  • CloudWatch
  • X-Ray
  • EC2

Frameworks

  • FastAPI
  • Flask
  • React
  • REST APIs
  • Mangum
  • Linux/Bash
  • CI/CD

Database

  • PostgreSQL
  • DynamoDB
  • SQLAlchemy
  • SQLite

Tools

  • Git
  • GitHub Actions
  • AWS SAM CLI
  • Docker
  • Docker Compose
  • Celery
  • Redis
  • Nginx
  • Pytest
  • Stripe
  • OpenAI API
  • Google Gemini API

About me

8th Grade

The beginning of my programming journey was in 8th grade — a Discord bot written in JS following a tutorial. The bots in the discord fasinated me and I wanted to make my own, but with actual code from scratch. It was simple and broke constantly. The codebase was a developer's nightmare, but it was my first real experience with programming and I loved watching my bot respond to commands in real time.

10th–11th Grade

By 10th and 11th grade I'd moved to C# and Unity, trading small bot scripts for simple mini games — engines, physics, state, all of it. That's really where I learned to think in systems without knowing it. Instead of the old simple tutorials, I mostly was recreating simple child hood games under the name Space High Gaming. The name sparked from my love for space and games.

After Graduation

after graduation I got serious about programming and wanted a professional career in it. I taught myself everything from there. No bootcamp, no CS degree — just documentation, broken project deployments, and enough courage to fix every annoying mistake or bug in the code, most of the time even staying up all night at 6AM debugging. Python became home for this since the reason I got into backend was to better my AI Engineering foundation later. which was funny because back in school, in my computer science class I thought Python was a joke and I would never use it because I was coding in C# at the time. I was very much wrong. Then portfolio projects to AWS: Lambda, API Gateway, DynamoDB, SQS, KMS. I passed the AWS Certified Developer Associate and Cloud Practitioner exams on the first attempt, and eventually I shipped a fix upstream to SlowAPI, the rate-limiting library a lot of FastAPI projects depend on. All within months of being self-taught and a passion to build.

Now

Currently I run RaynerCodes Cloud Solutions — I've built a live booking and payment platform for a real paying client, a fintech loan lending project API, apparently my own portfolio website (which your looking at now), and my own B2B SaaS product, LeverageCut. I'm looking for a professional backend or cloud engineering role, with healthcare and fintech at the top of my list, because in those environments the backend actually has to be right, not just functional, which is something I priortize when building systems. Cybersecurity and AI engineering are very interesting to me and they are the skills to follow.

Software (SaaS)

LeverageCut

Live Demo

Competitor price monitoring for e-commerce teams. Add a product URL, LeverageCut checks it on a schedule and flags the moment a competitor undercuts you — so you find out from a dashboard, not a lost sale.

  • Python
  • FastAPI
  • AWS Lambda
  • AWS SQS
  • DynamoDB
  • Framer
Visit site

RaynerCodes Cloud Solutions

Debo's Boxing and Fitness

Live Demo

A full-stack serverless booking and payment platform for a real paying client, with Stripe-integrated checkout, webhook-driven order confirmation, and automated conditional refunds. Solved a real double-booking race condition with an atomic conditional-write claim in DynamoDB, and layered in progressive brute-force lockout and IP-based rate limiting on the admin login.

  • Python
  • FastAPI
  • AWS Lambda
  • DynamoDB
  • API Gateway
  • Stripe
Visit site GitHub Repository

Personal Projects

Fintech Loan Lending Platform

A serverless loan lending API with credit-score based auto-decisioning, a one-way loan lifecycle state machine, and six separate customer-managed KMS keys to keep a security incident in one area from touching another.

  • Python
  • AWS Lambda
  • API Gateway
  • DynamoDB
  • SQS FIFO
  • KMS
  • CloudFormation
API Docs GitHub Repository

Looking for older work? View archived projects →

Get in touch

Hiring, collaborating, or just want to talk through something you're building — send it over.