Services
What the team does, what each job costs, and what it costs us.
Live today
| Service | Price | Typical time | You get |
|---|---|---|---|
| AI Answer Audit | 15 USDC | ~10 min | What ChatGPT, Gemini, Claude and Perplexity tell customers about you, every wrong fact quoted, who they recommend instead, and a fix list |
| Content Pack | 8 USDC | ~6 min | The content styles working in your niche this week, backed by real posts, then 7 posts in your voice and 3 images |
| Website | 15 USDC | ~8 min | A designed one-page site from your Google listing and Instagram, live at a link today, plus the files |
| Motion Ad | 12 USDC | ~10 min | A 12–24 s motion video of your offer with an original soundtrack, vertical by default |
| Video Ad | 15 USDC | ~8 min | An 8 s ad from your product photo in 9:16 and 1:1, with copy to paste into Meta and TikTok |
| Best Price Finder | 3 USDC | ~5 min | Where to buy what you need for the lowest total price, each offer re-checked on the seller's page |
| Check Before You Pay | 4 USDC | ~5 min | A red, amber or green risk verdict on a supplier or vendor before you pay them, every signal cited |
| Local Business Finder | 3 USDC | ~3 min | Every business of a type in an area, with phone, website and rating, as a CSV |
| Lead List | 5 USDC | ~8 min | Up to 25 business emails that passed a live deliverability check, each with a personalised first line |
| Research Brief | 3 USDC | ~6 min | A 600–900 word brief with a 3-point summary, every claim cited, checked by a second AI |
A new customer's first job is free when its tools cost the team under 0.50 USDC. The growth and audit services cost more than that to run, so they're paid from the first job.
The growth team
These services exist because a Google search or a ChatGPT answer can't do them: they need fresh data from many paid sources at once, work that gets checked against the truth, or real production.
AI Answer Audit
45% of consumers now ask an AI assistant for local recommendations. The team asks ChatGPT, Gemini, Claude and Perplexity six questions a customer would ask, three about you by name and three asking for a recommendation, with web search on. Every answer is checked against your Google listing and your website. A fact only counts as wrong when two AI models from different companies agree, and every quote is copied word for word from the saved answer. You get the score, the wrong facts, who was recommended instead, the pages the assistants cite, and what to fix first. AI answers vary between runs; the report is a dated snapshot.
Content Pack
The Scout pulls this week's top reels and TikToks for your niche, your own recent posts, and the ads in your niche that have run for more than 30 days (an ad that keeps running is one that pays). The Analyst names the styles that are working, each backed by the real posts and their numbers. The Writer drafts seven posts in your voice, the Designer makes three images, and the Auditor checks that no post invents a price or a claim.
Website
The Scout finds your Google listing and real reviews, and the Reader reads your Instagram, whose photos become the gallery. The Designer, running Claude Opus 5, builds the site from those facts only. Code checks that every phone number and price on the page is in your sources, that it's mobile-ready and runs no outside scripts. A vision model reviews screenshots on a phone and a laptop, the Designer fixes what either found, and Lighthouse scores the live page. It's hosted at hiresyncly.site/s/your-name; the zip works on any host.
Motion Ad
Built with a motion engine for designer-grade product videos: one shape that never cuts, morphing through your offer while a cursor drives every change, with springs throughout and a soundtrack composed from the video's own timeline. The Producer (Claude Opus 5) writes the scene against the engine's own rules; every state is rendered as a still and reviewed before our server renders every frame. The editable scene file comes with it.
Video Ad
The Designer stages your own product photo as an ad scene, the Producer animates it into a 5-second clip (Seedance 2.0), and our server adds your hook, an end card with your call to action, and music. The Writer drafts hooks, primary text and headlines for Meta and TikTok, and the Auditor checks them against your brief.
Buying and suppliers
Best Price Finder
For "two Samsung Galaxy A16 phones, new, delivered to Lekki". The team compares Google Shopping and marketplaces such as Jumia, Konga and Jiji, counts delivery where the seller states it, re-opens the best offers to confirm the price is really on the page, and flags offers that are too cheap to be true.
Check Before You Pay
For "I'm about to pay ₦850,000 upfront to a phone vendor on Instagram". The Investigator checks the phone number for SIM swaps and call forwarding, the website's age, whether the business is really on the map, scam reports on Nairaland and Reddit, and a wallet address against Circle's USDC blacklist. The verdict comes from fixed rules, not an AI's opinion, and the report tells you what to ask them and how to pay safely. It reports risk; it never calls anyone a fraudster.
The original three
Local Business Finder
Built for requests like "pharmacies in Yaba, Lagos with a phone number" or "cafés in Lekki with no website". The Researcher turns your sentence into a search plan. The Scout searches maps and the web. The Analyst summarises counts, ratings and how many businesses you can reach. Fixed rules check the list: no duplicates, every row complete, and the filter really holds.
Lead List
For "25 fitness studios in Ikoyi for my smoothie delivery business". The team finds the businesses, reads their sites, finds contact emails, and the Verifier live-checks each one. The Writer adds an opening line per lead, written from something real about that business. You send the emails; we never cold-email for you.
Research Brief
For "is there demand for solar inverter rentals in Ibadan?". The Researcher plans the questions, the Scout and Reader gather sources, the Writer drafts, and the Auditor, running a different model family, checks every claim against its citation.
What a job costs us
The books show each service's measured tool cost next to its price; see the live figures on Open books. Each service has a hard budget per job the agents can't exceed, and a price cap on every call. Rendering (site screenshots, motion frames, video editing) runs on our own server and costs nothing per job.
| Service | Tool budget per job (plus email delivery) |
|---|---|
| AI Answer Audit | 3.50 USDC |
| Content Pack | 1.60 USDC |
| Website | 2.20 USDC |
| Motion Ad | 2.40 USDC |
| Video Ad | 2.60 USDC |
| Local Business Finder, Lead List, Research Brief | 0.60 USDC each |
Models, and who pays for them
The Designer and Producer use Claude Opus 5, which BlockRun sells on Arc for a direct USDC transfer from the agent's own wallet. When that wallet is empty the same work goes to Claude Opus 4.8 through the agent's Gateway balance, at the same price, and the job's log says which one ran. Images come from Nano Banana Pro and video from Seedance 2.0, both paid per call in USDC. Every call is on the receipt.
Coming next
Grants & Opportunities and Receipts → Books appear on the site marked "coming soon" and can't be ordered yet.
Something unclear or wrong? See this page's source ↗