16 worked guides: which capabilities to call, in what order, what each call costs and what comes back. Every one runs on the same key and the same prepaid balance.
Agents: every guide is raw markdown at /use-cases/llms.txt.
New · GTM agent Your next GTM hire is a prompt. Find your audience, research their companies and draft your next campaign. One key. Pay per call. Watch the workflow → New · Introducing SEO for AI Agents Ask your agent to find the opportunity. It pulls live keywords, rankings, competitors, backlinks and page issues, then gives you the plan, not another dashboard. Read the launch → Introducing LinkedIn Automation for AI Agents Tell your agent who to find. It searches LinkedIn, enriches every match and starts personalized LinkedIn and email outreach, through one key and one balance. Read the launch → Introducing Claude for Cold Email Infrastructure Your agent can now buy the sending domain, wire SPF, DKIM and DMARC, create the inbox, send, and read what comes back. One balance, priced per call. Read the launch → Independent GTM playbook We built the first outbound workflow for Balance's founding GTM hire 100 live UK finance vacancies scanned, one hiring signal per account, qualification logic and hiring-trigger outreach. An executed GTM experiment. Read the playbook →200 competitor commenters in, 54 decision makers with a current work email out, and a person reads 36 profiles instead of 200. The comment is the trigger, never the opener.
Read the guide →social.postssocial.commentshistoryperson.enrichcontact.findcontact.verifylinkedin.inviteemail.sendBuild the list, verify it, and send from a domain that lands.
Companies that just posted the role you sell to, the person who owns the budget, and their verified email. Around two cents per finished contact.
The thing enrichment platforms charge a platform fee to orchestrate, as one call with one price and no credits to pre-buy.
The four calls behind a CRM row that is actually actionable, and the cache policy that keeps them from being expensive.
Three public, dated signals that separate accounts that are moving from accounts that merely fit.
The whole sending stack as API calls: domain, DNS, authentication, inbox, send, and the replies coming back.
Keywords worth ranking for, the SERP that owns them, and the pages behind it.
Two scraping providers behind one call, failover on every error, and one charge per page: the router turns a failed scrape into a second attempt instead of a broken step.
Search, read and cite the web from an agent with two calls, priced per search and per page, with the raw Google SERP as a third when you need rankings rather than sources.
Keywords worth ranking for, the SERP that owns each one today, and the gap against the competitors actually in your results.
One competitor domain in, their whole organic footprint and the keyword gap against yours out.
What ChatGPT, Claude, Gemini and Perplexity say when someone asks about your category.
The creatives competitors run, and the ones you ship next.
The creatives competitors run, diffed weekly, so you see what appeared, what died and what has been working long enough to copy.
Generate the ad, edit it, cut the product out, upscale for placement, then turn the winner into video.
Places, ratings and review text for local GTM.
Every provider's own API is addressable too. Ask /v1/discover in plain language and it returns capabilities and raw endpoints ranked for your job. Free, never runs a paid call.
Social and community
Public posts, profiles and comments as a research surface.
Turn competitor commenters on LinkedIn into a qualified lead list
200 competitor commenters in, 54 decision makers with a current work email out, and a person reads 36 profiles instead of 200. The comment is the trigger, never the opener.
Brand monitoring across web, social, news, reviews and AI answers
Every public mention of a brand, product, founder or competitor in the last 90 days, including the AI answers that name you, for a few dollars a sweep.
Social profile, post and comment data (X, Reddit, LinkedIn, TikTok, Instagram, YouTube)
Public profiles and recent posts for personalization that survives contact with a reader, plus the keyword search that finds people with the problem.