26+ years AI Automation · CV/OCR · Enterprise Systems Flutter · Python · .NET · TypeScript Production-first builder

Michael Russo

Senior Software Engineer — AI Automation, Computer Vision/OCR, Flutter Mobile, .NET & TypeScript

I design and ship end-to-end systems that connect real businesses to modern AI: cross-platform apps, secure APIs, streaming UX, CRM automation, and computer-vision/OCR document pipelines. Comfortable modernizing legacy platforms without breaking operations.

~$1.3M/yr DiSC Mobile platform revenue
AI in Production RAG, agent skills, real-time streaming
Enterprise-Ready Security, integrations, reliability

Explore: DiSC Mobile · AI tooling + projects · Technical skills

Technical Skills

End-to-end capabilities across AI, mobile, cloud, and enterprise systems

Core Programming Languages

Primary languages used in production environments across 26+ years.

TypeScript Python C# Dart VB / VB.NET JAVA C / C++ JavaScript SQL HTML/CSS

AI, Machine Learning, Computer Vision & OCR

LLM integration, RAG pipelines, embeddings, document intelligence, and applied computer vision.

Agentic ChatGPT / OpenAI Claude Gemini LM Studio Ollama MCP PyTorch TensorFlow Scikit-Learn OpenCV Tesseract / docTR OCR

Mobile & Client Apps

Cross-platform mobile development and desktop applications.

Flutter / Dart Android iOS Xamarin .NET MAUI .NET Blazor / Web Assembly Windows Forms WPF

Backend, APIs & Real-Time

High-performance backends with real-time streaming and secure auth.

Node.js / Express React ASP.NET Core .NET Signal-R REST WebSockets Socket Programming

Cloud, Data & Infrastructure

AWS and Azure cloud architectures, serverless systems, managed databases, authentication, storage, messaging, and containerized deployments.

AWS Azure Supabase PostgreSQL SQL Server SQLAnywhere Redis MongoDB DynamoDB Docker

Project Showcase

Deep dives into select projects — from AI-powered scraping to computer vision pipelines

⬡ Stage 1

Detect — Blueprint Object Detection

Streamlit demo detecting countertops, sinks, and appliances in architectural blueprints via trained Mask R-CNN model.

Blueprint detection: bounding boxes around detected countertops
Blueprint detection: pool house plan with detected surfaces
◈ Stage 2

Scale — OCR Measurement Extraction

Extracting blueprint scale via OCR (docTR) to compute real-world dimensions, areas, and generate material cost estimates.

Scale measurement: OCR extracting blueprint scale and generating cost table
Scale measurement: highlighted objects with dimensional data
◉ Stage 3

Segment — Instance Segmentation

Pixel-perfect instance segmentation masks isolating exact countertop shapes for area computation and material estimation.

Instance segmentation: green masks showing exact countertop boundaries

Model: Mask R-CNN with ResNet-50-FPN backbone · Training: 10,000+ annotated blueprint images · Infra: Trained via Riverpod on large GPU clusters · Demo: Streamlit interactive frontend

ScraperAI Platform
⚙ Stage 1

Dashboard & AI Provider Setup

Central dashboard for managing AI providers, prompts, scraping jobs, API keys, and analytics. Configure providers like OpenRouter and LM Studio with model selection and API key management.

ScraperAI dashboard with cards for AI Providers, Prompts, Scraping Jobs, Analytics, API Keys, and Settings
AI Providers screen showing Qwen Free (OpenRouter) and Qwen Coder (LM Studio) cards
Edit Provider modal with fields for Provider Type, Name, Description, API Key, and Default Model
✎ Stage 2

Prompt Engineering & Configuration

Create and manage scraping prompts with target URLs, custom instructions, system prompts, provider/model selection, output format (JSON/XML), and fetch method (headless browser or direct).

Prompts table listing Bobcat and West Herr prompts with Qwen Coder and Qwen Free providers
Edit Prompt form with Target URL, Prompt Body, System Prompt, Provider selection, Output Format, and Fetch Method
▶ Stage 3

Execution, Monitoring & Results

Monitor crawling jobs with status tracking, filtering, and pagination. View detailed job results including AI processing metrics (tokens, cost, confidence) and validated structured JSON output.

Crawling Jobs table showing 6 jobs with status badges, URLs, timestamps, and duration metrics
Job Details overview showing Target Information, Timing, Execution Status, and AI Processing cards
Analysis Results tab showing validated JSON output with vin_number and stock_number fields

Backend: .NET 8 / ASP.NET Core · Frontend: React SPA · Scraping: Headless browser (Playwright) · AI: OpenRouter, LM Studio, Anthropic, OpenAI providers · Output: Structured JSON/XML with validation

MathMCP Server
∑ The Challenge

Complex DSP Concepts

Excerpt from a Digital Signal Processing textbook. The material is dense, mathematical, and difficult to parse for beginners or high school students.

Textbook page with complex DSP equations and graphs
✓ The Solution

AI Simplification (LM Studio)

MathMCP server connected to LM Studio local LLM. The agent ingests the textbook image, simplifies the concepts for a high school level, and generates practice questions.

LM Studio chat window showing AI explanation of the math concept

Tech: Model Context Protocol (MCP) · Model: Local LLM (LM Studio) · Input: Visual / OCR + Text · Goal: Educational simplification & accessibility

Work & Impact

Key roles and the systems I've built

W-Interactive — Founder / Co-Owner, Lead Architect

Built and maintained enterprise mobility + document intelligence systems with long-term customer support, integrations, and modernization.

  • End-to-end ownership across mobile, backend, integrations, and automation.
  • Modernized legacy client apps/services while preserving uptime and workflows.
  • Delivered OCR/CV pipelines and automated POD generation in production.
Flutter C# / .NET Python OCR SQL Anywhere

Basil & "Annie" AI Sales Automation — AI Automation Architect

Conributed to AI-driven CRM automation platform for automotive dealerships: lead capture, qualification, follow-up, and CRM actions.

  • Built chat widget capturing leads and parsing VIN context on vehicle pages.
  • Implemented Node.js service routing leads into VinSolutions workflows.
  • Developed TypeScript agent skills for SMS/email/CRM operations.

InCommercial AI Chat Agent — Backend Security & Streaming

Security hardening + real-time AI streaming + RAG integrations for commercial real estate.

  • JWT auth + refresh token workflows and reliability improvements.
  • Helped polish front-end chat agent

Dialer + CRM Integrations — Real-Time Communication Systems

In-house dialing infrastructure and CRM automation integrations.

  • Custom socket-based dialing server (C#/.NET/WPF).
  • ScreenPop plugin for HubSpot and Redline CRM integrations.
  • Chat widget + lead capture + SMS/email automation workflows.

Projects & Innovation

Personal projects, open-source tools, and creative experiments

SmartCoder — AI Developer MCP Server

MCP server connecting AI coding agents to local environments

Model Context Protocol server for semantic code search with embeddings, automated file editing + analysis, and OCR-based visual inspection workflows. Designed for local LLM setups like LM Studio.

PythonFastMCPChromaDB SentenceTransformersOpenCVPyTesseract

Countertop (BlueprintCV)

Computer Vision for architectural blueprint analysis

Object detection + OCR scale extraction system trained on 10,000+ blueprint images. Detects countertops and appliances with a Streamlit demonstration frontend.

PythonPyTorchOpenCV docTRStreamlit
View showcase →

Task3r — AI Task Management Platform

Conversational AI task ops with serverless AWS backend

AI-powered task management for solopreneurs: chatbot interface, cloud-native serverless backend, companion marketing website.

FlutterAWS LambdaAPI Gateway CognitoSNSBedrockS3
Visit task3r.com →

ScraperAI Platform

AI-assisted web scraping with structured output

Configurable scraping system supporting local models or cloud LLMs, outputting structured JSON/XML data. Used for VIN extraction from dealership listings.

.NET 8ASP.NET Core ReactLLM APIs
View showcase →

MathMCP

Math education MCP server

MCP server for math education: converts complex text into simplified explanations, generates practice problems, interprets notation.

PythonMCP
View showcase →

Color Fluff

Mobile photo-to-coloring-book converter

Mobile application converting photos into coloring book or line-art images using on-device image processing.

FlutterDart OpenCV

Insurance Mobile App

Asset tracking & claim reporting

Mobile app allowing users to track valuables and generate insurance claim reports with secure cloud storage.

FlutterSupabase

Steward Mobile

Field service management platform

Service industry mobile platform for house cleaners and field services, connected to enterprise ASP.NET backend systems.

FlutterASP.NET Core

Let's Work Together

Whether you're looking for a senior engineer, AI automation expertise, or help modernizing a legacy system — I'd love to hear about it. Use the form or reach out directly.

Note: form endpoint will be wired to AWS API Gateway + Lambda when deployed.