GUID Solutions
GUID Solutions is Ronnie Guidry’s product studio for agentic software development, rapid prototyping, systems architecture, and decision systems. The analytics and experimentation background still matters, but the current work is focused on building intelligent products, operational tools, and working software with real technical depth.
Studio signal
15+ years across product analytics, experimentation, attribution, measurement, and business-to-technical translation now applied to AI desktop products, trading systems, research infrastructure, and full-stack software.
Current focus
What GUID Solutions Does
The studio is built around one practical question: what needs to exist, in software, for an idea to stop being abstract? In practice that means agentic software development, rapid prototyping, systems architecture, and decision systems across AI-driven desktop workflows, operational tooling, and full-stack products.
Translate ambiguity
Work usually starts with incomplete requirements, fuzzy workflows, or a strong intuition that has not been operationalized yet. GUID Solutions turns that into concrete product surfaces, data flows, and implementation choices.
Prototype with intent
The goal is not a slide deck or a speculative architecture diagram. It is a working slice: something people can use, react to, pressure-test, and refine into a more durable product.
Engineer for iteration
That includes local and cloud AI workflows, desktop and web delivery, operational controls, observability, and data-informed iteration without turning every decision into an analytics exercise.
Operating model
The common thread is technical product building under real constraints: translating intent into interfaces, choosing the right architectural boundary, and getting to usable software without losing rigor.
Featured Products
Current work is organized around products rather than generic service categories. Each one reflects a different slice of the same studio model: combine product judgment, technical depth, and disciplined iteration until the system is actually useful.
AI-powered desktop product
An AI-powered desktop product for homeschooling families and elementary educators who need print-ready K-6 worksheets, lesson plans, and answer keys without hand-assembling materials from generic chatbot output.
Teacher’s Assistant is the most architecturally substantial product in the current portfolio because it has to feel like software, not a thin prompt wrapper. The workflow is guided, stateful, and oriented around artifacts that are actually usable in a classroom or homeschool setting.
The desktop shell is built with Tauri v2, with a React + TypeScript frontend and Zustand-managed application state driving guided creation flows, previews, export controls, and local filesystem integration. The creation experience is designed around wizards and structured inputs so users can move from idea to lesson packet with less prompt-writing friction.
Behind that interface is a Node/Express generation API that streams progress over SSE, orchestrates prompts, parses inspiration material, and produces HTML-first educational artifacts that can move cleanly into print and PDF/export workflows. Supabase handles authentication, projects, credits, and version history, while generation can run through premium cloud models or local Ollama-based workflows depending on the user’s environment and budget.
System design
Why it matters
Technology
Intraday day trading bot
An intraday day trading bot that turns validated research into real-time paper and live execution workflows with explicit operational controls.
Alpaca is deployed on GCP and structured around dedicated test and production runtime environments, controlled releases, and close operational oversight. It is designed as a real system with distinct runtime responsibilities rather than a monolithic trading script.
The runtime handles real-time market-data streaming, strategy routing, risk-based sizing, and execution controls using contract-driven interfaces that keep downstream behavior predictable as strategies evolve. An OpenAI copilot sidecar is used in shadow and test-oriented review workflows to help inspect behavior without turning autonomous review into production authority.
Runtime design
Operational posture
Technology
Quantitative research platform
A deterministic quantitative research and validation platform for strategy discovery, built to complement Alpaca by deciding what should reach runtime in the first place.
Forrest is the research and validation layer that sits upstream of Alpaca. It focuses on disciplined discovery rather than curve-fit storytelling, with regime-aware analysis and subgroup discovery used to pressure-test where a strategy idea appears to hold up and where it does not.
Walk-forward validation, multiple-comparison guardrails, and deterministic export contracts are central to the system. The point is to create a reproducible research path that can promote strategies into downstream execution systems without hand-wavy translation.
Research design
Role in the system
Technology
Collaboration product
A full-stack collaboration product for recurring groups and meetup coordination, designed to make lightweight community operations less chaotic.
Helm is built as a React + Express + PostgreSQL/Drizzle application with the basics treated seriously: authentication, membership management, scheduling, and shared notes. The aim is not just event scheduling, but a better operational memory for recurring groups.
AI-assisted import and enrichment workflows help turn messy source material into structured knowledge that a group can actually reuse over time. That makes Helm as much about continuity and capture as it is about scheduling.
Application design
Product angle
Technology
Background
Ronnie Guidry spent more than 15 years working across product analytics, web measurement, attribution, experimentation, funnel optimization, and the translation layer between business context and technical execution.
That experience still shapes the work. It sharpens instrumentation choices, keeps experimentation grounded in decision quality, and helps separate signal from noise when products start interacting with real users and operational constraints.
But the current center of gravity is product building: AI-enabled software, working prototypes, decision systems, and architecture that supports actual use instead of abstract intent.
Working Style
The studio works best where product strategy and implementation are tightly coupled. That means prototyping early, letting architecture inform scope, and using data discipline to improve decisions rather than delay them.
Prototype-first
The first milestone should usually be a working product slice, not an inflated roadmap. That creates real feedback and exposes technical constraints before they become expensive.
Architecture matters
Decisions about contracts, state, local versus cloud execution, and operational visibility change what a product can become. They are not backend housekeeping.
Use data well
Metrics, instrumentation, and structured validation still matter. They just serve the product instead of becoming the product.
Ship the real thing
The standard is software that can be used, reviewed, and improved. If a decision should stay abstract, it probably is not ready to dominate the roadmap.
Contact
GUID Solutions is open to collaboration where product thinking and technical depth both matter. Use the form for a direct note or connect through LinkedIn.