12
AI Agent

AI Resume Screener
for Small Business

Small businesses are drowning in resume emails — reading PDFs one by one, copying notes into a spreadsheet. This ends that. Every resume that hits Gmail is scored in seconds, logged automatically, and — if it clears 60 — surfaced in Slack with a hiring decision already drafted. Change roles by editing three sentences of plain English. Running cost: under $5/month.

Result6hr manual review → 45 minutes, shortlist ready
How it works
1Gmail trigger fires on every incoming resume email
2PDF extracted from attachment and sent to the AI agent for a full read
3AI scores the candidate 0–100 against role criteria written in plain English
4Every candidate logged to Google Sheets — no one falls through the cracks
5Score ≥ 60 triggers a Slack message: tier, 3 reasons, red flags, next step
Gmail Trigger PDF Extraction AI Scoring Google Sheets Slack Hiring Automation AI Agent n8n
14
AI Agent

Auto Email Bot —
Inbox That Replies Itself

Replies are where leads go to die — sitting unread while you're busy. This bot watches the inbox around the clock, reads every incoming reply, classifies the intent, and fires back a contextual response — no human in the loop. Interested leads get a meeting link. Questions get answered. Not interested gets a graceful close. Every thread handled, every time.

How it works
1Trigger fires the moment a new reply lands in the inbox
2AI reads the full thread and classifies intent — interested, objecting, asking a question, or unsubscribing
3A context-aware reply is drafted and sent automatically within seconds
4CRM is updated with the reply status so no lead falls through the cracks
Inbox Monitoring AI Reply Classification Auto-Response Intent Detection CRM Update n8n
01
AI Agent

Automated Business
Research Agent

What takes a sales rep two hours per prospect — Google searches, LinkedIn tabs, scattered notes — takes this agent 90 seconds. Drop in a company name, get back a structured intelligence brief in your inbox: key people, recent activity, and red flags. Ready before the call even starts.

How it works
1Triggered by a company name via webhook or form input
2Agent searches Google, LinkedIn, and Wikipedia in parallel
3AI structures and compiles a full research brief
4Formatted report sent to Gmail automatically
AI Agent Google Search API LinkedIn Scraping Wikipedia API Gmail API Automated Reporting
02
RAG & LLMs

PDF Intelligence
with Pinecone RAG

A 200-page supplier contract. A policy manual nobody reads. A compliance report from three years ago. This system lets you have a natural conversation with any document — ask in plain English, get precise answers with source references. Built for businesses drowning in PDFs they can never actually search.

How it works
1PDF uploaded and split into semantic chunks
2Chunks embedded and stored in Pinecone vector DB
3User asks a question in plain language
4Claude retrieves relevant context and answers precisely
RAG Pipeline Pinecone Claude API Vector Embeddings PDF Processing
03
Content Automation

AI Content Brief
Generator

Briefing a writer used to mean an hour of research, a messy doc, and a guess. This system takes a keyword, analyses search intent, identifies the angle competitors are missing, and outputs a complete SEO brief in under a minute — headings, talking points, audience, and word count guide included.

How it works
1Topic or keyword passed as input
2LLM chain researches context and search intent
3Structured brief generated with headings and talking points
4Brief exported as a ready-to-use document
LLM Chain Content Strategy SEO Automation Prompt Engineering
04
Content Automation

RSS Feed to
Notion Database

Staying ahead of your industry shouldn't mean tab-switching between ten blogs every morning. This system monitors curated sources around the clock, pulls every new article the moment it publishes, summarises it with AI, and files it as a structured record in Notion — your competitive intelligence feed, fully automated.

How it works
1Scheduler polls 10 RSS feeds at regular intervals
2New articles are detected, deduplicated, and parsed
3AI generates a concise summary for each article
4Structured record — title, source, summary, link — saved to Notion
Content Automation RSS Parsing Notion API AI Summarisation Content Aggregation
05
AI Agent

Intelligent Email
Classifier

The average business owner loses two hours a day to inbox triage. This agent reads every incoming Gmail, understands its intent — lead enquiry, support issue, invoice, or noise — and routes it to the right category instantly. High-priority emails get an immediate alert. Everything else is sorted before you open your laptop.

How it works
1Triggered instantly when a new email lands in Gmail
2AI reads the subject, sender, and body to determine intent
3Email is labelled and routed into the correct category
4High-priority emails trigger an immediate alert notification
AI Agent Gmail API NLP Classification Auto-Labelling Inbox Automation
06
Monitoring Agent

Automated Website
Ping Tester

Every minute your site is down costs you visitors who won't come back. This agent pings your URLs every 5 minutes, around the clock — no subscription, no monthly fee. The moment a page returns an error or goes dark, you get an email alert before your users notice anything is wrong.

How it works
1Scheduler triggers a ping to each monitored URL every 5 minutes
2HTTP status code is captured and logged for every request
3Any non-200 response or timeout flags the site as down
4Instant email alert fired with the site URL and status code
Monitoring Agent HTTP Status Checks Uptime Monitoring Gmail API Scheduled Automation
09
AI Agent

AI Content Approval
Pipeline

AI writes fast. Humans catch what AI misses. This pipeline keeps both in the loop — Claude drafts platform-specific posts, surfaces them in Slack for a one-click review, then either publishes automatically or loops your feedback back for a rewrite. Human-in-the-loop content, without the bottleneck.

How it works
1Claude drafts Twitter and LinkedIn posts for a given topic
2Draft is sent to Slack with a link to an approval form
3Human approves or rejects with feedback via n8n form
4Approved posts publish to Twitter and LinkedIn automatically
5Rejected posts are rewritten by Claude and loop back for review
HITL Claude AI Slack Integration n8n Forms Twitter API LinkedIn API Feedback Loop
10
Multi-Agent

Manager Agent +
Specialist Sub-Agents

One AI agent hits its limits fast. This system uses a Manager AI to receive a task and delegate intelligently — spinning up a Researcher, a Writer, and a Formatter as specialist sub-agents, each running as its own sub-workflow. The manager chains their outputs and returns a polished final result. The architecture scales to any number of specialists.

How it works
1User sends a task via chat trigger to the Manager Agent
2Manager calls the Researcher sub-agent to gather facts on the topic
3Research output is passed to the Writer sub-agent to create content
4Written content is passed to the Formatter sub-agent for clean markdown
5Manager returns the final formatted result to the user
Multi-Agent Claude AI Execute Workflow Agent Hierarchy Task Routing Sub-Workflows n8n
11
Content Automation

AI Social Media
Content Machine

Posting consistently across three platforms shouldn't require a content team. This pipeline generates fresh ideas, writes platform-specific copy for X, Facebook, and Instagram, renders a pixel-perfect image from an HTML template, and schedules everything through Buffer — daily, without a single manual step.

How it works
1Scheduler triggers daily — AI generates fresh content ideas for the niche
2Claude writes tailored copy for X, Facebook, and Instagram separately
3HTML template rendered into a pixel-perfect Instagram image via HTML-to-JPG
4All posts + image sent to Buffer — scheduled and queued automatically
Content Automation Buffer API Claude AI HTML to Image Instagram Facebook X / Twitter n8n Scheduled Posts
15
AI Agent

Law Firm AI Intake —
4 Stages, Zero Manual Work

From form submission to a fully-populated Clio matter — signed, paid, and filed — without a single person lifting a finger. Claude scores every lead in under 10 seconds, drafts the engagement letter, triggers DocuSign, fires the LawPay link the moment the signature lands, and notifies the attorney on Slack before the first call. No attorney time. No paralegal. No follow-up chasing. Ever.

Result3hr manual intake → 4 minutes, automated
How it works
1Prospect submits form — Claude reads the case and scores it 1–10 in under 10 seconds. Score ≥ 6 proceeds; lower scores get a polite rejection automatically
2Claude drafts a complete engagement letter — scope, fees, retainer terms, client obligations — and sends it via DocuSign for e-signature on their phone
3Signature confirmed → PCI-compliant LawPay payment link fires instantly. Payment clears → receipt sent → Stage 4 triggers automatically
4A fully-populated Clio matter is created — case type, AI score, retainer amount, signed date. Attorney gets a Slack alert and email summary before the first call
Claude AI AI Lead Scoring DocuSign LawPay Clio Slack Gmail n8n Law Firm Automation Client Intake
I also build for: Law Firms Dental Practices Hire an n8n Expert Personal Trainers Travel Agents
Want Something Like This?

Let's build a system
tailored to your business.

Every project above started with a 30-minute call. Tell me your bottleneck — I'll design the solution.

Reach out directly

Let’s talk about your
automation needs.

Pick any channel below — I usually reply within a few hours.