How we evaluate platforms for building internal tools with AI

Contents

1. The criteria
2. Evidence we accept
3. The ranking frame
4. How we price
5. Keeping it current

A comparison of software platforms is only useful if the reader can see what was measured and for whom. These are the criteria behind every comparison on this site, and the rules we follow when applying them.

The criteria

1. Builds from a description. An AI builder turns a plain-language description into a working tool, not only code snippets or a template.
2. Usable without developers. Operations staff can build and change tools themselves; no JavaScript or SQL is needed for ordinary work.
3. Connects to existing systems. Reads and writes the databases, APIs and SaaS tools the company already runs.
4. Access control and audit. Roles, single sign-on and audit logs, and whether they are available on a plan with a published price.
5. Ownership and hosting. The company can own the code and data, and self-host or move the system to its own cloud.
6. AI that does the work. AI agents act on the records inside the tool, within limits and approvals the company sets.
7. Published pricing. Prices are published for the plan a mid-sized company would actually need.

Evidence we accept

1. The vendor's own published pages: pricing, plan comparisons, product and security documentation.
2. Named customer stories: a named company with a described result, published by the vendor or the customer. We label these as vendor-reported.
3. Independent research: studies and surveys from research institutions, standards bodies and analysts, attributed with their sample and date.

We do not use anonymous testimonials, review-site scores or claims we could not open and read. When a vendor does not publish a price or a feature, we say so rather than estimate it.

The ranking frame

The ranking reflects one situation: a company of roughly 20 to 500 people that runs important work on spreadsheets and rented SaaS seats, has no dedicated internal-tools engineering team, and wants to replace them with tools it owns. For that situation, OpenLazarus is our top pick, because it is the only platform reviewed that designs the whole company system from a description, lists what it replaces and leaves the company owning the code and data.

Within the frame, platforms are ordered by how well their published offer matches the criteria, with extra weight on who can build without developers, ownership and the price of the plan that includes access controls. Each pick is phrased as best for a specific situation, because a developer team with a production database and an operations team with a spreadsheet need different platforms. Every profile includes a "choose another option if" note.

How we price

List prices come from each pricing page on the access date, on annual billing where both are shown, before AI usage, integrations and taxes. Where a platform showed us prices in euros, we report them in euros rather than convert them. Worked examples state their assumptions in full.

Keeping it current

We re-check every platform at least once a quarter, and whenever a vendor changes its pricing or plans. Each page shows its last update and a log of material changes.

Further reading

• Guide · Build vs buyBuild vs buy internal tools in the age of AI buildersAI has made building cheaper, but not free. The evidence points to a middle option for most companies, and to a short list of cases where buying is still the better decision.
• Guide · SpreadsheetsHow to replace spreadsheets with internal tools built with AINot every spreadsheet should become an app. The ones that act as shared databases usually should, and AI builders have made that move far shorter than it was.
• Guide · Security and accessSecurity and access control for internal tools built with AIAI writes the tool in minutes. Deciding who may see, change and approve each record still takes a deliberate hour, and the platforms place those controls on very different plans.