Consulting and operations
Email triage and follow-up system
Consulting and operations workflow
Problem
Important emails, leads, and follow-ups get buried in inboxes, leading to missed opportunities and inconsistent response times.
What was built
I built a structured email system that classifies incoming messages, identifies priority items, and supports follow-up through tracking and draft generation.
Result
The system gives clearer visibility on what needs action, reduces missed follow-ups, and speeds up response handling without relying on memory.
Workflow pattern
Incoming email -> classification -> priority tagging -> tracking -> follow-up timing -> draft reply -> human review.
Review boundary
All outbound communication remains human-reviewed before sending.
Tools used
- Gmail
- AI classification layer
- Google Sheets
- Automation workflows
What this shows
Inbox management improves when decision-making is structured, not when more tools are added.
Where this applies: This same pattern fits service businesses where important client emails, leads, or follow-ups are buried in a shared inbox or handled from memory.
Consulting pipeline
Lead tracking and follow-up system
Consulting pipeline workflow
Problem
Lead research, outreach, and follow-up are often scattered across spreadsheets, inboxes, and notes, making it easy to lose track of warm prospects.
What was built
I built a lightweight system combining lead capture, outreach tracking, and follow-up scheduling in a single workflow using existing tools.
Result
The system made follow-up more consistent, improved pipeline visibility, and reduced reliance on memory or ad-hoc tracking.
Workflow pattern
Lead source -> business details -> outreach -> follow-up schedule -> next action tracking.
Review boundary
Messaging decisions and follow-up strategy remain human-controlled.
Tools used
- Google Maps lead sourcing
- Google Sheets
- Gmail
What this shows
The improvement came from workflow clarity, not introducing a heavy CRM.
Where this applies: This same pattern fits service businesses managing enquiries, quotes, warm leads, or outbound prospects without a structured pipeline.
Sales and communication
AI-assisted follow-up and reply drafting system
Sales and communication workflow
Problem
Writing follow-ups and replies repeatedly is time-consuming and often inconsistent across conversations.
What was built
I built a system that generates context-aware draft replies using previous email threads and structured workflow triggers.
Result
The system reduces drafting time and makes communication more consistent without fully automating responses.
Workflow pattern
Tracked conversation -> trigger -> retrieve context -> generate draft -> human review -> send.
Review boundary
All drafts are reviewed and approved before sending.
Tools used
- Gmail
- AI drafting layer
- Workflow automation
- Tracking system
What this shows
AI is most useful when it supports communication, not replaces it.
Where this applies: This same pattern fits teams that repeatedly write follow-ups, enquiry responses, client updates, or internal status messages.
Content operations
Content and SEO workflow system
Content operations workflow
Problem
Content creation is inconsistent when it starts from a blank page without structure, research, or clear intent.
What was built
I built a workflow connecting keyword intent, topic clusters, research notes, and structured briefs before drafting content.
Result
The system creates more consistent output, stronger alignment with SEO intent, and easier reuse of research and ideas.
Workflow pattern
Topic idea -> keyword intent -> research -> content brief -> draft -> internal linking.
Review boundary
Final positioning, tone, and publication decisions remain human-led.
Tools used
- Local wiki
- Keyword mapping
- Content briefs
- Web research
- Static content system
What this shows
AI writing becomes significantly more useful when attached to a structured editorial process.
Where this applies: This same pattern fits businesses that need repeatable proposals, reports, client updates, knowledge articles, or internal documentation.
Internal knowledge
Knowledge capture and retrieval system
Internal knowledge workflow
Problem
Useful ideas, research, and insights are often lost across bookmarks, notes, and scattered documents.
What was built
I built a local knowledge system that captures, organises, and connects information to ongoing projects and workflows.
Result
The system creates a growing knowledge base that supports writing, consulting work, and product development over time.
Workflow pattern
Capture -> organise -> connect to projects -> retrieve -> reuse.
Review boundary
The system supports thinking but does not replace judgement or decision-making.
Tools used
- Obsidian
- Structured notes
- Project-linked documentation
What this shows
Knowledge systems only work when they are tied to real work, not just storage.
Where this applies: This same pattern fits small teams where process knowledge, client context, or training notes are scattered across inboxes, documents, and individual memory.
AI product systems
AI product systems: Embodifi and Persuade Write
Product development workflows
Problem
AI products require more than model output. They need structured workflows, safety layers, and usable interfaces.
What was built
I built two production-grade AI systems: a storytelling engine that turns user input into structured narrative outputs, and an email analysis and rewriting system with scoring and improvement suggestions.
Result
The systems demonstrate how structured workflows improve reliability, usability, and user control in AI products.
Workflow pattern
User input -> interpretation -> structured generation -> validation -> output.
Review boundary
Outputs are guided by defined system rules, with user control over final use.
Tools used
- LLM APIs
- Structured pipelines
- Frontend interfaces
- Backend workflows
What this shows
Useful AI systems require orchestration, not just prompts.
Where this applies: This same pattern applies when a business needs AI to work reliably inside a product, internal tool, or repeated operational workflow.
