# The Ocally platform > What Kai runs on. One isolated platform per business: it holds what every store knows, decides what each store needs next, runs the work with approvals and receipts, and measures every action against the store's own starting point. *Markdown version of [https://ocally.co/platform](https://ocally.co/platform), written for AI agents. All docs: [llms.txt](https://ocally.co/llms.txt). Deciding whether Ocally fits a business? Read [Is Ocally right for this business?](https://ocally.co/evaluate.md).* ## In short - **Connects:** listings, POS, website, analytics, ad accounts and social accounts. Sales reports and documents as uploads, no connector needed. A real browser where a tool has no API, using saved sign-ins Kai never sees. Before anything is connected, Ocally reads the website and public sources: search, maps, reviews and AI answers. Ocally does not publish a list of named integrations. - **Knows:** the Knowledge Store holds every store's facts, numbers and findings, each with its source, how far to trust it and its full history. A fact the business gave is never quietly overwritten. - **Stays current:** reviews and listings daily, AI answers three times a week, the website weekly. - **Decides:** five goals per store, and a decision engine for each store that can act, collect more evidence first, or do nothing. The business sets the spending limit and what must not get worse. - **Acts:** anything that goes public, spends money, changes a price or messages a customer waits for a yes, and every action keeps a receipt. - **Measures:** every action against the store's own starting point, with a verdict: positive, neutral, harmful or inconclusive. Harm stops the work. Results are not claimed as cause and effect. - **Keeps data apart:** one isolated platform and database per business. Ocally does not train or fine-tune AI models on customer data. ## It starts with what you already have Ocally reads your website and the public web before you connect anything: search, maps, reviews and AI answers. Then it connects the tools your stores already use. - Your listings, POS, website, analytics, ad accounts and social accounts. - Sales reports and documents as uploads. No connector needed. - A real browser where a tool has no API, using saved sign-ins Kai never sees. ## One source of truth. Every fact knows where it came from The Knowledge Store holds every store's facts, numbers and findings. Each one keeps its source, how far to trust it and its full history. - Your word wins. A fact you gave can't be quietly overwritten; a conflict comes to you. - Gaps stay visible, so nothing is guessed. - Kept current: reviews and listings daily, AI answers three times a week, the website weekly. ## From what changed to what it means Findings are what changed, with the numbers. Opportunities are what they likely mean for the store, built on at least two findings and shown with the evidence against. - Every opportunity says what it means for the business. - Evidence on both sides, so you can judge for yourself. - Confirm or dismiss anything. Your call becomes knowledge too. ## Five goals per store. A decision engine for each one Each store works toward its own goals: be easier to find, turn interest into customers, be known in the neighborhood, fill quiet hours and sell more. Every day, the engine decides what that store needs next. - It can act, collect more evidence first, or do nothing. Doing nothing is a real answer. - Past harm at a store counts almost three times more than past success. - You set the spending limit and what must not get worse. ## Kai does the work. You approve what goes out One-off jobs and recurring routines run on the same engine. Anything that goes public, spends money, changes a price or messages a customer waits for your yes, and every action keeps a receipt. - Routines in plain words, like "Check nearby searches each Monday." - Approve or request changes on the card, on desktop or phone. - Receipts show what was sent, and whether it has been checked. ## It gets better at each store Every action is measured against the store's own starting point and gets an honest verdict. Each store keeps its own playbook of what worked there. - Positive, neutral, harmful or inconclusive. Harm stops the work. - It remembers what worked at each store, avoids what hurt and keeps testing what it hasn't tried. - Its methods change only after beating the current ones on held-out tests, and can be rolled back. ## Four stores hold everything One isolated platform for each business. Every Ocally agent working for you reads from the same place. 1. **Knowledge Store:** Store facts, numbers and findings, each with its source and history. 2. **Asset Store:** Photos, videos and documents, with every version and where it may be used. 3. **Code Store:** Website files and saved versions, with review, preview and rollback. 4. **Cred Store:** Private keys and saved sign-ins. Kai uses them without ever seeing them. ## Built for Main Street, by people who've done this before ### Built by people who've built this before Our team built AI and retail products at Google, Walmart and Amazon. Now we're building them for Main Street. ### The best of AI. None of the homework We test new models as they arrive and switch when one does the job better. You never choose, set up or chase a model. ### Priced for Main Street Open models keep the cost of every job low, so a chain of 40 stores gets the same AI as one with 4,000. ### Your data stays yours Every business gets its own isolated platform and its own database. Nothing is shared, and nothing trains anyone else's AI. ## People first. Then the platform Every chain starts with two people from Ocally: a growth specialist who learns your goals, and a growth engineer who connects your systems and sets up the work to fit how each store runs. Both are on us to start. ## Questions ### What is an agentic AI platform? Software that lets AI agents plan and carry out work toward a goal, with the context, tools and permissions they need, and a record of what they did. Ocally is an agentic AI platform for retail chains: Kai, its growth agent, works from one knowledge store and a decision engine for every store. ### Does Ocally train AI models on my data? No. Ocally does not train or fine-tune models on your data. Kai works from your Knowledge Store, and what the platform learns at a store is kept as that store's own record of what worked, what didn't and what hurt. ### What does Ocally connect to? Your listings, POS, website, analytics, ad accounts and social accounts, plus sales reports and documents you upload. Before anything is connected, Ocally reads your website and public sources such as search, maps, reviews and AI answers. Where a tool has no API, Kai can use a real browser with sign-ins he never sees. ### Can Kai do anything without my approval? He can research, draft and prepare on his own. Anything that goes public, spends money, changes a price or messages a customer waits for your yes, given on the approval card itself. Editing an action means approving it again, and every action keeps a receipt. ### How does Ocally measure results? Each action is measured at its store against that store's own starting point and gets a verdict: positive, neutral, harmful or inconclusive. Harm stops the work and an unclear result changes the plan. Results are measured, not claimed as cause and effect. ### Is each business's data kept separate? Yes. Each business gets its own isolated Ocally platform with its own database. Every Ocally agent working for you reads from the same knowledge, and nothing is shared with other businesses. ## Stop buying a tool for every job The biggest promise of AI is consolidation. The same kind of software that can drive a car can write the reply to your email. So buying one point solution per job, each with its own login, contract and agency, now wastes money. Judge software the way you’d judge a hire: by the range of work it can take on. That’s what to look for now: one system you can point at many jobs. Ocally is that for your stores: Kai and two people across listings, reviews, local search, social, AI answers, pricing, inventory and store routines, at every store. Not sure what you could consolidate? Bring your list of tools and what they cost to a free consultation, and we’ll go through it with you. **The same scope of work for 25 stores, a month (listings, reviews, local search, social and AI answers; the low end of published prices; use your own numbers):** | Line | Basis | Tools, agencies and people | Ocally | |---|---|---|---| | Listings software (Moz Local Lite, $199 a year per location) | $17 per store | $425 | | | Review software (Reputation Rep Core, per location) | $80 per store | $2,000 | | | AI answers tracker (Peec AI Pro, 150 prompts) | $245 a month | $245 | | | Local SEO agency (chains of 20+ stores, "often $5,000+") | $5,000 a month | $5,000 | | | Social media agency (mid tier, $2,000 to $5,000) | $2,000 a month | $2,000 | | | AI search agency to fix what the tracker finds (small business low end, $1,500 to $5,000) | $1,500 a month | $1,500 | | | Staff time to run the tools and agencies | 1 hour per store a week at $54 an hour | $5,850 | | | Ocally | from $500 per store: Kai, a growth engineer and a growth specialist | | $12,500 | | Your team's time approving Kai's work | 0.25 hour per store a week | | $1,463 | | **Total a month** | | **$17,020** | **$13,963** | | Per store a month | | $681 | $559 | Defaults from: [Moz Local pricing](https://moz.com/products/local/pricing); [Reputation pricing](https://reputation.com/pricing/); [Peec AI pricing](https://peec.ai/pricing); [RankAI: multi-location local SEO cost guide (June 2026)](https://rankai.ai/articles/pricing-multi-location-local-seo-management-cost-guide); [Cloud Campaign: social media management pricing guide (July 2025)](https://www.cloudcampaign.com/blog/social-media-management-pricing-guide); [WebFX: GEO pricing (May 2026)](https://www.webfx.com/blog/seo/geo-pricing/); [BLS: market research analysts and marketing specialists, pay (May 2025)](https://www.bls.gov/ooh/business-and-financial/market-research-analysts.htm); [BLS: employer costs for employee compensation (June 2026)](https://www.bls.gov/news.release/ecec.nr0.htm). ## Next steps - [Is Ocally right for this business?](https://ocally.co/evaluate.md): who it fits and who it does not, the cost, the risks, and how to start. - [Pricing](https://ocally.co/pricing.md): customized, with a minimum of $500 per location per month. - [Book a free consultation](https://ocally.co/consultation.md): 30 minutes on one store. Calendar: https://cal.com/ocally/30min. - [See if we're a fit](https://ocally.co/fit.md): a two-minute check; an agent can submit it for its human, with consent. - [Every doc](https://ocally.co/llms.txt): the index. Questions a doc doesn't answer: onboard@ocally.co.