Open research desk / Catalog 001
Evidence before
endorsement.
A practical resource center for people choosing AI building tools. Define the job, collect evidence, price the whole workflow, and keep an exit path.
Start with sequence
Do not begin with a logo wall.
The right resource depends on the job, evidence, data path, reviewer, operating cost, and exit. The center keeps those questions visible before any commercial shortlist.
- 01Open guide ↗
Define the job before choosing the tool
Turn a broad AI idea into one observable task, owner, input, output, and review point.
- 02Open guide ↗
Run a one-hour evidence check
Replace feature-list browsing with a short, documented proof using your own representative examples.
- 03Open guide ↗
Map data and model access
Trace what enters the system, who can see it, what is retained, and which boundary must be verified.
- 04Open guide ↗
Price a repeatable workflow
Estimate the full operating cost of a workflow, including review, exceptions, integration, and exit work.
Six resource types
Compare the operating shape, not the category label.
AI-assisted coding environments
Editors, agents, and review systems that help people design, write, test, and maintain software.
Evaluation dossier ↗TYPE 02Model playgrounds and APIs
Interfaces for testing models, structured output, tool use, retrieval, and production calls.
Evaluation dossier ↗TYPE 03Workflow automation platforms
Tools that move data and actions between systems through triggers, steps, branches, and approvals.
Evaluation dossier ↗TYPE 04Evaluation and observability tools
Systems for test cases, traces, quality review, cost inspection, regressions, and operating feedback.
Evaluation dossier ↗TYPE 05Knowledge and retrieval systems
Pipelines and stores that help an AI workflow find approved organizational knowledge.
Evaluation dossier ↗TYPE 06Deployment and data platforms
Hosting, storage, identity, queues, and operational services that turn a build into a maintained product.
Evaluation dossier ↗Search the open catalog
Find the next useful question.
19 records shown. Search runs only in this browser.
Guide
Define the job before choosing the tool
Turn a broad AI idea into one observable task, owner, input, output, and review point.Guide
Run a one-hour evidence check
Replace feature-list browsing with a short, documented proof using your own representative examples.Guide
Map data and model access
Trace what enters the system, who can see it, what is retained, and which boundary must be verified.Guide
Price a repeatable workflow
Estimate the full operating cost of a workflow, including review, exceptions, integration, and exit work.Guide
Audit export and exit options
Know what can be exported, in which format, and what must be rebuilt before the tool becomes essential.Guide
Plan human review and escalation
Define who checks the work, which failures stop the process, and where judgment cannot be delegated.Guide
Move from prototype to operating tool
Separate a convincing demonstration from a workflow that can be owned, monitored, corrected, and stopped.Resource type
AI-assisted coding environments
Editors, agents, and review systems that help people design, write, test, and maintain software.Resource type
Model playgrounds and APIs
Interfaces for testing models, structured output, tool use, retrieval, and production calls.Resource type
Workflow automation platforms
Tools that move data and actions between systems through triggers, steps, branches, and approvals.Resource type
Evaluation and observability tools
Systems for test cases, traces, quality review, cost inspection, regressions, and operating feedback.Resource type
Knowledge and retrieval systems
Pipelines and stores that help an AI workflow find approved organizational knowledge.Resource type
Deployment and data platforms
Hosting, storage, identity, queues, and operational services that turn a build into a maintained product.Field kit
AI job brief
A one-page worksheet for defining the task before evaluating a tool.Field kit
Evidence log
A repeatable record for inputs, configurations, outputs, corrections, and decisions.Field kit
Total workflow cost sheet
A worksheet that combines vendor cost with human and operating work.Field kit
Data path map
A worksheet for tracing data classes, processors, retention, people, and approved use.Field kit
Review and escalation plan
A compact operating control for review roles, rejection reasons, stops, and escalation.Field kit
Exit rehearsal
A checklist for exporting, archiving, rebuilding, and safely ending a tool dependency.See the work in context
Tools make more sense beside a real build.
Made by AI documents finished projects, the people accountable for them, the stack, the wrong turns, and what stayed human.
Explore Made by AI builds ↗