WEB · AI · INTEGRATIONS

Practical problems. Purpose-built software.

I’m Ricky Criss, a developer studying AI software engineering. I bring 20 years of hands-on experience in the metal industry to building websites, workflow integrations, and custom applications.

Open to remote developer opportunities & scoped freelance projects

WORK WITH ME

A clear problem. A useful solution.

Start with a focused scope and an outcome we can test together. These are proposed service packages; scope, timing, and price are agreed after discovery.

Website launch & care

For a small business that needs a clear, professional place to explain its work and help customers take the next step.

  • Responsive service pages and clear calls to action
  • Contact flow, accessibility basics, and launch checks
  • Source handoff and a simple update guide

Package: A scoped launch, with optional ongoing updates and care.

Discuss a website

Connect tools. Reduce busywork.

For a team copying the same information between forms, spreadsheets, or business tools.

  • Map one workflow and its source of truth
  • Connect agreed systems through supported APIs
  • Test duplicate records, failures, and recovery

Package: Discovery followed by a bounded implementation; optional maintenance.

Discuss a workflow

Have a different technical challenge?

I also explore AI assistants, custom applications, backend systems, and remote hardware support. We’ll establish feasibility and a realistic scope before making commitments.

THE PERSON BEHIND THE BUILD

A practical builder,
always learning.

Ricky Criss / Criss Tech Solutions

My background is in the metal industry, where I spent 20 years working with real materials, constraints, and problems. That experience shapes how I approach software: understand the job, break it into manageable parts, and build something useful.

I’m currently studying AI software engineering and developing projects across web applications, AI integrations, and backend systems. I also build and configure computers, from workstations to gaming systems.

I’m looking for remote opportunities where I can contribute, learn from others, and take responsibility for a defined piece of work.

Explore my GitHub

PORTFOLIO REPOSITORY

Built to learn. Shared to demonstrate.

Full project index with architecture deep-dives, live prototypes, and repository links.

ProEval Studio AI Code and RLHF Evaluation Simulator
Featured Case Study AI & RLHF React 19 Gemini API Pyodide WASM

ProEval Studio

An interactive AI code-evaluation and RLHF sparring workbench. Prepares engineers for Outlier AI and Code4Ai screenings with timed bug-hunt sprints, in-browser Python WASM execution, A/B model auditing, and Gemini-powered adversarial QA reviews.

Key Architecture: Gemini Structured Outputs (responseSchema), Pyodide sandbox, Firebase Auth & anti-tamper Firestore security.

Vivid-Architect Autonomous Unreal Engine 5 Developer Agent on Poe
Active Poe Agent Unreal Engine 5 LoRA SFT 57MB RAG

Vivid-Architect (vivid_ue5)

An autonomous Unreal Engine 5 C++ developer agent powered by fine-tuned Qwen2.5-Coder-1.5B with an indexed 57MB Epic Games engine corpus. Scaffolds paired headers/sources, validates Hungarian notation, and injects Build.cs modules—deployed via FastAPI Poe bridge.

Key Architecture: LoRA Rank-16 on RTX 3070 Ti, 1,619 engine headers RAG index, AST C++ validator, Poe server bot.

Mortal Online 2 Vista
Leaflet Firestore AI Assistant

The MO:2 AI Assistant

An interactive mapping project for Mortal Online 2, bringing location information into a navigable map interface. Built around Leaflet and Firestore to explore interactive interfaces and shared data.

Skills explored: Mapping UI, application state, and data integration.

Corner14
React AI Vite

Corner14

A React and Vite application developed with Google AI Studio. A learning project exploring how to structure an AI-facing application and manage interactive interface state.

Skills explored: React components, state management, and AI application interfaces.

Custom ReAct Agent
Personal project AI Implementation

Tool-calling AI agent

A Python ReAct agent example using the Gemini SDK to explore tool calling and multi-step tasks. A starting point for understanding how an assistant can choose tools, act, and respond.

Skills explored: Python, Gemini SDK, tool calling, and agent control flow.

HOW WE’D WORK TOGETHER

Know what happens next.

  1. 01

    Understand

    Define the problem, users, tools, and what a useful result looks like.

  2. 02

    Agree

    Set deliverables, exclusions, milestones, access, and price before implementation.

  3. 03

    Build & test

    Work in small steps, share progress, and check the agreed workflow.

  4. 04

    Hand over

    Document the result and discuss any support or improvements you need.

Have a project—or a place on your team?

I’m open to remote developer opportunities and focused freelance projects. Share the problem you’re solving, the tools involved, and what you need next.

Connect on GitHub

GitHub is my public contact route for now. If we’re already connected on another platform, send the project link there. Please keep credentials and confidential material out of public issues.

J