Ten AI solutions β already built, deployed, and running across four live platforms. Not concepts: production systems ready to productise for clients.
Every inbound lead scored, enriched, and prioritised β before a human ever sees it.
The problem
Inbound leads sit unsorted in an inbox. Reps waste time on low-intent enquiries while a lead with a fresh funding round goes cold overnight.
What it does
Built with
Claude (Bedrock) Β· MCP tool servers Β· CRM integration
Proof it works
Runs live and unattended across InAgentic's own inbound pipeline.
Ideal customer
B2B sales teams drowning in inbound volume with no dedicated SDR
Idea to live, on-brand article in under three minutes β no copy-paste.
The problem
Content teams spend hours on research, drafting, editing and manual publishing for every single article.
What it does
Built with
Claude Β· LangChain Β· CMS integration (Ghost)
Proof it works
Powers the live blogs behind InAgentic.ai and FileDone today.
Ideal customer
Marketing teams and solo founders who need consistent content without a writer on payroll
From signup to fully provisioned member β with no human required.
The problem
New signups need manual account setup, access grants and welcome emails β a bottleneck that delays activation and burns admin time.
What it does
Built with
Claude Β· LangGraph Β· AWS SQS + App Runner Β· LangSmith
Proof it works
Provisions a new member end-to-end in seconds β running in production today.
Ideal customer
Membership sites, course platforms and SaaS products with a manual onboarding step
Topic in. Branded avatar video, published to YouTube. No editing software opened.
The problem
Producing short-form video β course lessons, marketing, social β usually needs a presenter, a camera, and hours of editing per clip.
What it does
Built with
Claude Β· HeyGen Β· Custom MCP servers Β· YouTube API
Proof it works
Used to produce InAgentic's own course and marketing video content.
Ideal customer
Course creators, coaches and marketing teams who want a face on video without filming
Free answers for browsers. AI-powered conversion for signed-in visitors.
The problem
Generic chatbots either cost too much to run at scale, or never actually convert a visitor into a booking.
What it does
Built with
Claude Sonnet (Bedrock) Β· Google Calendar API Β· pgvector RAG search
Proof it works
Live on InAgentic.ai; reused on dogrun.ai with zero new infrastructure.
Ideal customer
Service businesses and course sites that want a chatbot which books revenue, not just answers questions
Photo or email in. Structured document out.
The problem
Important content β curricula, meeting notes, plans β gets scattered across whiteboard photos and phone notes, and never becomes a usable document.
What it does
Built with
Claude (Bedrock) Β· Gmail integration
Proof it works
Used to turn whiteboard sessions into InAgentic's own course curriculum.
Ideal customer
Trainers, consultants and small teams who capture ideas on the fly
The same AI that runs your chat β now answering your phone.
The problem
Phone enquiries still need a human to answer, book or follow up β and most small businesses have no after-hours cover.
What it does
Built with
Claude (Bedrock) Β· Vapi Β· Twilio Β· WhatsApp Business API
Proof it works
Added as a new channel with zero duplicated infrastructure β same calendar, same model, same number.
Ideal customer
Any business that loses enquiries to voicemail outside office hours
A conversational guide through compliance paperwork β filed correctly, first time.
The problem
Company filings and regulatory paperwork eat founder time, and carry real financial risk when filed late or wrong.
What it does
Built with
Claude API Β· Companies House / HMRC integration Β· GOV.UK Design System
Proof it works
MVP built by an engineer with hands-on delivery on DEFRA's medicine and water-abstraction licensing systems.
Ideal customer
UK startup founders and small-business owners without an in-house company secretary
One dashboard, one codebase, every client's data kept apart.
The problem
Agencies running the same AI tooling for multiple clients usually end up maintaining a separate codebase β and a separate cost β per client.
What it does
Built with
Next.js Β· AWS Amplify Β· Shared Postgres Β· Role-based access
Proof it works
Runs four live brands today on one shared backend for roughly Β£18-28/month in infrastructure.
Ideal customer
Agencies and AI consultancies selling the same automation to many clients
Constraints aren't restrictions β they're what makes an agent trustworthy enough to deploy.
The problem
Fewer than 1 in 4 companies using AI have a working governance framework, and EU AI Act fines can reach 3% of global turnover.
What it does
Built with
LangGraph interrupt() Β· Postgres checkpointing Β· ISO/IEC 42001
Proof it works
Same approval-gate pattern already runs inside InAgentic's own production pipelines.
Ideal customer
Any regulated or risk-conscious business deploying AI agents that take real-world actions
Every product here is already live in production β not a concept. Tell us which one fits and we'll scope it.