// product engineer · full stack · builder
Robert Stumborg.
I build things that work.
Full stack product engineer. I've spent 15 years building the kind of software other businesses run on; SaaS platforms, mobile apps, ML pipelines, operational tools. I work best when I own the full surface: architecture through deployment.
Languages
- Python
- PHP
- JavaScript
- TypeScript
- SQL
Backend
- FastAPI
- Django
- Symfony
- Laravel
- Express.js
- PostgreSQL
Frontend
- React
- React Native
- Next.js
- Angular
- Vue.js
Infrastructure
- Docker
- CI/CD
- Linux
- Bash
- Git
- GitHub Actions
- REST APIs
Product & Design
- Figma
- Illustrator
- Wireframing
- Conversion Optimization
- Growth Marketing
EventJetter is a multi-tenant B2B SaaS platform for event travel agencies. Its centerpiece is event discovery: agents search any destination and instantly see every concert, game, and show happening within their trip dates — powered by a locally-synced catalog with fuzzy search, trending performers, and destination-level browsing. From there, they assemble complete travel packages with aligned hotels and flights.
The agent workflow covers itinerary building, client management, package lifecycle tracking (draft → live → archived), and line-item pricing. A standout feature is the automated marketing pipeline: Remotion renders branded video ads in multiple formats and durations, Playwright generates social media banners, and AI-assisted tools produce contextual captions and email copy — all from package data.
The core architecture challenge was clean multi-tenancy with three permission tiers (app admin → agency admin → agent). Doctrine query extensions automatically scope all data access by agency and role, so controllers and repositories never handle tenant filtering directly.
Integrates with event inventory, hotel, and flight providers, plus AI-assisted tools for marketing copy.
shooter.bet is a real-time match prediction platform where users wager virtual currency on individual rounds of live CS2 bot matches streamed on Twitch. Each round opens a betting window during freeze time, users pick a side, and payouts settle instantly when the round ends — a parimutuel pool model where winners split the losers' pot proportionally.
The core technical challenge was building a reliable real-time pipeline from game to browser. CS2's Game State Integration — configured on the game client, not the server — pushes match data to a Laravel backend webhook, which processes round transitions, manages betting windows, calculates payouts with row-level locking and atomic balance updates, and broadcasts everything to Angular clients through Laravel Reverb WebSockets. A priority-based job queue (Beanstalkd) ensures bet payouts are never delayed by lower-priority work.
A Python automation daemon bridges the platform to the physical CS2 server — controlling matches via RCON, managing the game client through xdotool, and orchestrating OBS scene transitions for the Twitch stream. A custom SourceMod plugin fetches player model data from the API to apply user-customized bot appearances in-game. The platform also runs AI-driven bot users with eight distinct betting personalities and context-aware chat, creating organic-feeling activity around the clock.
Features include multi-method auth (Discord, Steam, X/Twitter SSO), a 50-level XP/achievement system, map voting with currency-weighted votes, user-to-user money transfers, a referral program, and a custom async IRC client built on AmphPHP for alternative user verification.
- Restructured the platform's payment architecture to support multiple processors, then built one integration end-to-end and led the team through a second, including merchant boarding, webhooks, and fund distribution
- Refactored the transaction fee calculation engine and built an automated test suite that validates against a QA spreadsheet, achieving comprehensive coverage across billing configurations
- Developed and implemented a credits system for the platform (data model, API, transaction logic, and refund handling) integrated into the existing invoicing and payment pipeline
- Automated multiple internal billing processes, including transaction fee and recovery fee invoicing, replacing manual workflows with scheduled platform features that generate and deliver invoices automatically
- Conducted technical interviews for developer and QA candidates, mentored team members through code reviews, and delivered team presentations on automatic testing and AI workflows
- Built the platform from scratch — React Native mobile app, Laravel API, and a Python/PyTorch ML classification pipeline that normalized product data across retailers with different naming, categorization, and units
- Designed data visualization layer using d3.js to surface pricing trends and cross-retailer comparisons, turning raw scraped data into actionable insights for users
- Established data quality validation pipelines to ensure accuracy and completeness of ingested retailer pricing data across heterogeneous sources
- Engaged CostDetect as a client, building the initial platform architecture that grew into a full-time product role
- Managed full project lifecycles from client consultation and requirements gathering through deployment and ongoing maintenance across multiple concurrent engagements
- Helped scale a DTC consumer brand to $25k/day in revenue, managing product, affiliate partnerships, campaign performance, and margins across the full operation
- Built and maintained high-traffic ecommerce storefronts with shopping cart, payment processing, and product search functionality on UltraCart
- Designed marketing assets and product creatives in Illustrator, working across the full funnel from ad creative to landing page to checkout
- Built a dynamic quiz and landing page generator for ringtone campaigns, producing artist-specific funnels at scale for a client
- Ran independent affiliate marketing campaigns at scale, applying platform-building and high-volume email experience to generate revenue across multiple verticals
- Built a multi-tenant data exchange platform used by marketing companies to share and route lead data between networks -- effectively a B2B SaaS product before the term was mainstream
- Developed an internal platform for creating, managing, and sending email campaigns at scale using PowerMTA for high-volume delivery, used as core operational tooling across the business
- Collaborated with marketing and business development teams to translate campaign requirements into technical solutions, optimizing email deliverability and building reporting dashboards for performance tracking