Skip to main content

Data Decision Software

Data is the most Valuable Asset.

Use it to pick the data component for the job, then verify each tool against your data footprint before you adopt it.

Jobs to be Done​

Capabilities​

Products​

Pipeline Engineering

Agents​

Useful tools to convert documents into knowledge, taken from eliza framework:

  • folder2knowledge
  • knowledge2character
  • tweets2character

Example:

npx folder2knowledge <path/to/folder>
npx knowledge2character <character-file> <knowledge-file>

Failure Modes​

  • Adopting a tool before naming the decision job it serves — the catalog becomes a shopping list, not a choice.
  • Choosing an analytics or BI stack without checking who owns the underlying data.
  • Treating a link list as a recommendation instead of evidence that a category has demand.

Context​

  • Agent Development — build the agents that query and act on this data.
  • Data Flow — the Clean, Fast, Open test each data tool must pass.
  • Data Engineering — the pipeline layer beneath these toolkits.
  • Platform — where owned data capability becomes reusable.

Questions​

Which data toolkit component — ETL pipeline, semantic layer, or AI query interface — is most ready to be replaced by AI-native alternatives that don't require SQL or data engineering expertise?

  • At what data complexity level does a dedicated data toolkit become essential versus AI agents querying raw data directly?
  • How does the convergence of agent frameworks with data pipelines change the architecture for teams that currently maintain separate analytics and AI systems?
  • Which data toolkit bottleneck — ingestion latency, query performance, or model freshness — is most commonly the reason analytics projects stall before producing value?

Changes my mind: if AI agents reliably query raw data without a dedicated toolkit, the case for maintaining a separate analytics stack weakens.

Next question: which component in this toolkit should be owned rather than rented once a data footprint becomes strategic?