Personal WebsiteAustin Knapp
ResumeSay Hello
Hi, I'm Austin

AI, data, and product systems builder focused on real-world deployment

I build systems that turn ideas into real, working products.

70,000+ monthly visits across products I've built

800,000visits and growing

Projects

Selected Work

RentPad AI

Multi-tenant AI property management SaaS

Founder-led product build spanning architecture, onboarding, pricing, and the day-to-day workflows landlords and tenants actually need.

  • Built and architected a multi-tenant platform supporting landlord and tenant workflows
  • Designed onboarding and pricing; piloted with 8 landlords and property owners
  • Structured shared services for auth, billing, and notifications with Supabase and Stripe

Why It Matters

This is the clearest example of end-to-end ownership across product decisions, system design, and shipping a real SaaS experience.

TypeScriptNode.jsSupabaseRailwayStripe

Motor Intelligence

ML estimation + incentive forecasting automation

Production-facing forecasting and process automation work built around recurring OEM incentive estimation and monthly operational workflows.

  • Automated OEM incentive forecasting with a machine-learning estimation approach
  • Scaled monthly processing with Excel VBA automation pipelines integrated with cloud workflows
  • Built a repeatable monthly estimation and reporting workflow instead of one-off manual runs

Why It Matters

Shows applied ML in a business setting where the work has to be usable, repeatable, and reliable inside real operating constraints.

PythonExcel VBAForecastingData Pipelines

NexStratus

Early-warning ML system for healthcare supply-chain risk

Capstone early-warning system focused on turning healthcare supply-chain signals into earlier, more actionable risk visibility.

  • Detected supply-chain risk up to 72 hours in advance
  • Built an end-to-end analytics pipeline with feature engineering and model evaluation
  • Framed the system around earlier detection of disruption signals rather than retrospective reporting

Why It Matters

A strong end-to-end ML example: problem framing, feature engineering, evaluation, and output designed for actual decisions.

PythonXGBoostEvaluationGitHub

Work I Like

What I Build

AI + ML Systems

Designing practical forecasting and automation workflows that integrate into production operations.

PythonForecastingModel EvaluationXGBoost

Data Engineering

Building structured pipelines for ingestion, transformation, and repeatable reporting at scale.

SQLData PipelinesExcel VBAAutomation

SaaS Architecture

Shipping multi-tenant products with role-based workflows, onboarding, and monetization layers.

TypeScriptNode.jsSupabaseStripe

Analytics Delivery

Turning data into decision-ready insights through dashboards, KPI tracking, and business reporting.

TableauPower BIExcelAPIs

Quick Proof

Signal, not filler.

A few fast indicators of the kind of work I actually do: product architecture, production ML, real traffic, and a stack that spans backend, data, and delivery.

Four snapshots

Traffic

1,000,000+

annual page views supported across deployed work

Product

Multi-tenant SaaS

built from scratch with auth, workflows, onboarding, and billing

ML Delivery

Forecasting + automation

production pipelines designed for recurring operational use

Core Stack

Python / SQL / TypeScript / Node.js

comfortable across backend systems, data work, and product delivery

Execution

What That Looks Like In Practice

Capability

Multi-tenant SaaS Architecture

Evidence

RentPad AI: built and architected a multi-tenant platform for landlord and tenant workflows, plus onboarding and pricing design.

Stack

TypeScriptNode.jsSupabaseStripe

Capability

ML Forecasting Automation

Evidence

Motor Intelligence: automated OEM incentive forecasting with a machine-learning estimation approach.

Stack

PythonForecastingAutomation

Capability

Operational Pipeline Scaling

Evidence

Motor Intelligence: scaled monthly processing using Excel VBA automation pipelines integrated with cloud workflows.

Stack

Excel VBAData PipelinesCloud Workflows

Capability

End-to-End Analytics and Evaluation

Evidence

NexStratus: built feature engineering and model evaluation pipeline that detected supply-chain risk up to 72 hours in advance.

Stack

PythonXGBoostModel Evaluation

Timeline

Where I've Been Spending Time

  1. Motor Intelligence - Data and AI Insights Analyst

    May 2024 - Present

  2. RentPad AI - Co-Founder & CEO

    Nov 2024 - Present

  3. Sequoia Apps - Founder & Lead Developer

    Jul 2025 - Present

  4. NexStratus - Capstone Project Team Lead

    Oct 2025 - Present

  5. Feeding The NRV - Founding Member & Director of Digital Strategy

    Feb 2024 - Present

How I Operate

Tools + Working Style

Skill Matrix

Languages

PythonSQLJavaScriptTypeScript

Backend + Product

Node.jsAPIsMulti-tenant SaaSStripe

Data + ML

ForecastingFeature EngineeringModel EvaluationAutomation Pipelines

Platforms + Analytics

SupabaseRailwayLinuxGitHubTableauPower BIExcelVBA

How I Operate

  1. 01Design the architecture and data flow before implementation.
  2. 02Ship for production constraints: performance, reliability, and maintainability.
  3. 03Automate repetitive workflows and remove manual operational overhead.
  4. 04Tie technical output back to business impact and measurable outcomes.

FAQ

Questions People Usually Ask

How can I evaluate technical depth if the code is private?

You can review project architecture details here and request a walkthrough focused on design decisions, tradeoffs, and delivery constraints.

What kinds of roles are the best fit?

AI engineering, data engineering, operations automation engineer, and startup roles where technical ownership spans architecture through shipping.

What is the fastest way to reach out?

Email [email protected] with role context and timeline. LinkedIn outreach is also active.

Why are some project details intentionally limited?

Most production work is under NDA, so this portfolio emphasizes system scope, outcomes, and stack decisions without exposing private implementation.

Have something interesting?

Hiring, collaborating, or just want to talk about AI systems, product, or messy data problems that need to become real software? Email is the fastest way to reach me.

Email meDownload ResumeLinkedInGitHub

Most production code is private because a lot of it was shipped under NDA.