Jeremy Longshore

What I’m Building

From concept to production, here’s what keeps me up at night (in a good way).

Perception

Live at perception-with-intent.web.app | AI News Intelligence

Stop manually checking 50 news sources. Perception monitors everything, filters what matters using 8 specialized Vertex AI agents, and delivers executive-level insights to your dashboard.

Multi-agent system orchestrated via Google’s A2A Protocol on Vertex AI Agent Engine. Source-agnostic ingestion (RSS, APIs, custom connectors). Real-time Firebase dashboard with Gemini 2.0 Flash analysis.

Stack: Python 3.11, Firebase, Vertex AI Agent Engine, Gemini 2.0 Flash Features: 8-agent orchestration, daily executive briefs, smart alerts, enterprise security Google Cloud: Vertex AI, Cloud Run, Firebase Hosting, Firestore, Cloud Scheduler

Claude Code Plugins

Live at claudecodeplugins.io | 253 Plugins

The only Claude Code marketplace 100% compliant with Anthropic’s 2025 Skills schema. 253 production-ready plugins including 185 Agent Skills with tool permissions, version tracking, and enhanced activation triggers.

First marketplace to implement the complete 2025 specification released by Anthropic. Includes MCP servers, slash commands, subagents, and the Skills Powerkit - the first Agent Skills meta-plugin.

Stack: Next.js 15, TypeScript, GitHub-based marketplace, Node.js MCP servers Features: 253 plugins, 185 Agent Skills, tool permission system, activation guides Status: v1.3.1 (November 2025), 221+ commits in 2025

PipelinePilot

Production | B2B Sales Automation Platform

B2B sales automation orchestrating external data providers (Clay, Apollo, Clearbit, Crunchbase) through an AI-powered Reasoning Engine. Firebase Functions gateway routing to Vertex AI for intelligent API orchestration.

Serverless architecture with bring-your-own-keys (BYO) policy for cost control and compliance. Multi-cloud ready with Terraform templates for GCP, AWS, and Azure deployments.

Stack: Firebase Functions Gen2 (Node 20 ESM), Python ADK, Vertex AI Reasoning Engine Features: 4 ADK-compliant agents, 6 connector tools, action counting, multi-tenant isolation Google Cloud: Firebase Functions, Vertex AI, Firestore, Secret Manager

HustleStats

In Development | Youth Sports Analytics

Youth soccer statistics tracking with three integrated systems: core Next.js app, Vertex AI multi-agent orchestration (A2A protocol), and NWSL video pipeline using Veo 3.0.

Currently in Phase 1 migration from PostgreSQL/NextAuth to Firebase Auth and Firestore. Step 1 complete with local Firebase wiring verified. Multi-agent system includes orchestrator plus 4 sub-agents for operations automation.

Stack: Next.js 15, React 19, TypeScript, Firebase Auth, Firestore (migrating from Prisma) Features: Player profiles, game tracking, Vertex AI agents, NWSL video generation (CI-only) Google Cloud: Firebase Hosting, Cloud Functions, Firestore, Vertex AI Agent Engine, Veo 3.0

Start AI Tools

Live at startaitools.com | AI Implementation Hub

AI tools directory, tutorials, and implementation guides. Built in under 24 hours. 65+ posts, 344 pages, technical deep-dives on actually deploying AI.

Not another AI hype blog - practical guides on what works, what doesn’t, and how to ship fast.

Stack: Hugo, Archie theme, Netlify, Pagefind search Purpose: Help others deploy AI solutions in days, not months

DiagnosticPro

95% Production Ready | AI Diagnostic Platform

AI-powered equipment diagnostic platform. Upload a problem photo, get Vertex AI Gemini analysis for $4.99. Helps people avoid getting overcharged by repair shops.

API Gateway routes Stripe webhooks to private Cloud Run backend. Vertex AI analyzes images, generates PDF reports, delivers via email. Data infrastructure includes 266 BigQuery tables for diagnostic intelligence.

Stack: React, Express, Vertex AI Gemini, API Gateway, Cloud Run Features: Image analysis, PDF generation, Stripe integration, email delivery Google Cloud: API Gateway, Cloud Run, Firestore, Vertex AI, Firebase Storage, BigQuery Status: Awaiting Firebase Storage bucket creation (5-minute console setup)

Waygate MCP

GitHub | Enterprise MCP Server

Security-hardened Model Context Protocol server with Docker isolation. TaskWarrior integration, FastAPI backend, Nginx proxy. Enterprise-grade security without enterprise-grade complexity.

Stack: Python 3.12, FastAPI, Docker, Nginx, TaskWarrior Security: Read-only filesystem, non-root user, dropped capabilities Status: v2.1.0 stable

Bob’s Brain

GitHub | Sovereign AI Agent

Personal AI agent that runs wherever you want. Swap LLM providers like batteries (Claude, Google, OpenRouter, Ollama). Learns from every interaction. Optional Slack integration.

Built on the principle that AI should serve you, not surveil you.

Stack: Python, Flask, SQLite/Postgres, Neo4j, Redis, Chroma Philosophy: Your hardware, your rules, your data

Private AI Deployments (Intent Solutions)

intentsolutions.io/ai-models | Enterprise Private LLM Stacks

Keep data on your cloud while offering ChatGPT-style experience to every teammate. Curated model stack includes Llama 3.1 70B (enterprise reasoning), Mistral 7B (fast interactions), and Qwen 2.5 14B (technical/multilingual).

Orchestration layer automatically routes calls based on task requirements. Maps workflows during onboarding for consistent performance. Combined with Hybrid AI Stack for 60-80% cost reduction via intelligent local/cloud routing.

Stack: Llama 3.1 70B, Mistral 7B, Qwen 2.5 14B, Docker, Ollama, n8n, Terraform Features: Model orchestration, workflow mapping, local/cloud routing, AWS/GCP ready For: Enterprises wanting private AI without cloud API bills

N8N Workflow Automation

Production Workflows | 10+ Active

Enterprise automation without expensive platforms:

Stack: N8N, Claude API, various integrations Impact: Eliminating manual work, one workflow at a time

Prompts Intent Solutions

Live Catalog | 150+ Templates

Battle-tested prompts for development, business ops, marketing, finance. 74 specialized Claude Code agents included.

Stop starting from scratch. Use what actually works.

For: Developers and business users who value their time

This Site

jeremylongshore.com | Personal Blog

Personal blog running hugo-bearblog theme. Clean, minimal, fast. Auto-syncs select content from Start AI Tools daily.

Stack: Hugo, hugo-bearblog theme, Netlify, GitHub Actions Performance: Built and deployed in minutes

In Progress

Local RAG Agent — Retrieval-augmented generation that runs entirely local RSS Atoms — Feed validation system (226+ feeds tested and curated) AI DevOps — AI-powered DevOps workflows and automation

Open Source Philosophy

Most projects are MIT licensed on GitHub. Clone them, fork them, learn from them, build with them.

If it solved a real problem for me, maybe it’ll help you too.

Background

20+ years in restaurant operations (Bonefish Grill, Waffle House) → self-taught developer → AI automation specialist. Marine Corps Reservist, Citadel grad, now building systems that eliminate bullshit and save people money.

Google Cloud Platform Expert: Extensive experience with Vertex AI, BigQuery, Firestore, Cloud Functions, Cloud Storage, and GCP infrastructure. Built production systems processing millions of records on Google Cloud.

Based in Gulf Shores, Alabama. Available for consulting through Intent Solutions.

Contact: Email: jeremy@intentsolutions.io Discord: asphaltcowboy X: @asphaltcowb0y GitHub: @jeremylongshore LinkedIn: jeremylongshore