Pay per call
Subscription fatigue: paying for tools you use twice
Subscription fatigue in numbers: firms leave 36% of software licences unused on average, and UK consumers spend 1.6 billion pounds a year on unwanted subscriptions.
The IKAG blog
Plain-English guides to AI agents, the tools and data they can’t reach on their own, the subscriptions you pay for and barely use, and how to check what your AI tells you.
Written by Ronit Ahuja, founder of IKAG. RSS feed llms.txt
Pay per call
Subscription fatigue in numbers: firms leave 36% of software licences unused on average, and UK consumers spend 1.6 billion pounds a year on unwanted subscriptions.
Data your AI can call
How AI video is sold in 2026, from monthly credits to per-second API pricing, how to make 20 ad takes without a plan, and the quality, rights and labelling checks.
AI agents and tools
ChatGPT Dots, Grok Bot and Meta’s Muse are AI agents with their own cloud computers. What each one does, who can get it and what none of them can reach.
The subscriptions you pay for and barely use, and what paying per call changes.
Pay per call
Subscription fatigue in numbers: firms leave 36% of software licences unused on average, and UK consumers spend 1.6 billion pounds a year on unwanted subscriptions.
Pay per call
Not one per AI app: a single account at a service can work in ChatGPT, Claude and Cursor. What adds up is a subscription to each service your AI uses.
Pay per call
Monthly plans suit steady use and pay per call suits bursts. How they compare for AI tools: quotes before a call, spending limits, receipts and your own keys.
What today’s AI agents can do, how they connect to tools, and how to set them up.
AI agents and tools
ChatGPT Dots, Grok Bot and Meta’s Muse are AI agents with their own cloud computers. What each one does, who can get it and what none of them can reach.
AI agents and tools
Step by step: add a custom connector or MCP server in ChatGPT, Claude, Gemini, Cursor, VS Code, Claude Code and Codex, with plan limits as of October 2026.
AI agents and tools
MCP, the Model Context Protocol, lets ChatGPT, Claude, Gemini and other AI apps use outside tools. What it is, how connecting works and how to stay safe.
AI agents and tools
Two ways to connect AI agents to outside data: a router that offers many paid tools behind one key, and a data network that returns the record behind each value.
AI agents and tools
Search, page extraction, specialist data APIs, generation models and MCP: how an AI agent reaches the outside world, with a worked example on real records.
AI agents and tools
Eight questions for any agent data gateway, applied to IKAG in early access: provenance, failures, routing, cost, own keys, retention, terms.
AI agents and tools
What IKAG, in early access, means by an intelligent data network for AI agents, and how it compares with single APIs, marketplaces and MCP servers.
Leads, AI video, local businesses, social media and the web, without a plan for each.
Data your AI can call
How AI video is sold in 2026, from monthly credits to per-second API pricing, how to make 20 ad takes without a plan, and the quality, rights and labelling checks.
Data your AI can call
Where an AI agent’s list of businesses and decision makers comes from, why models alone get work emails wrong, the UK and US email rules, and what a good list includes.
Data your AI can call
How an AI agent can list local businesses with ratings, reviews, websites and opening hours: map platforms, open data and business websites, their terms and freshness.
Where an answer comes from, what its source record holds, and how to check it.
Answers you can check
Pew Research found Google users clicked a result in 8% of visits with an AI summary and 15% without. What that means for answers you need to check.
Answers you can check
What a Legal Entity Identifier is, how an AI agent should query the GLEIF register, and why “no match” or “lapsed” never means a company is gone.
Answers you can check
A practical guide for agents and developers: what LEIs, SEC CIKs and Wikidata IDs identify, who issues them, how to resolve each, and when to link them.
Answers you can check
Seven parts that make an AI answer checkable, from the source and record ID to attestation dates and unknowns, shown on seven real company lookups.
Answers you can check
A five-step check for any AI answer, worked through on real public records, and what good provenance looks like when you need to rely on a fact.
Answers you can check
Seven offline requests to IKAG’s first capability, in early access: what it answers, what it leaves unknown, and why a shared name never joins two records.
Early access
IKAG is opening to early users and data partners first. Join the early-access list and we’ll email you when your place opens.
Get early access