Independent Project

S-Curve Data

Building toward an AI product data science team.

I'm building S-Curve Data to explore how specialist AI agents can work together—from understanding product performance and predicting outcomes to measuring impact and informing the next decision. The work combines methodological rigor with shared evidence, traceable assumptions, and human judgment.

The team vision

What I'm building toward

My experience in product data science has taught me that a useful analysis starts with a clear question and ends with a decision someone can act on. I'm designing the Autonomous Data Team (ADT) to connect those steps: specialists working from shared metric definitions, carrying evidence between workflows, and keeping consequential decisions under human review.

The architecture brings analytics and data quality together with forecasting, predictive modeling, experimentation, and causal inference. Planning and orchestration would coordinate the work, while insights and memory would carry prior findings and unresolved questions into the next cycle.

  1. 01 · Plan

    Frame the product question, define success, and make forecasts, scenarios, and assumptions explicit.

  2. 02 · Prove

    Test claims with the right methods, diagnostics, and validation, keeping uncertainty visible.

  3. 03 · Report

    Bring sources, findings, assumptions, and caveats together into evidence people can inspect.

  4. 04 · Decide & learn

    People review the tradeoffs and choose actions. Outcomes and open questions inform the next cycle.

Causal Inference and Forecasting are live today. Predictive Modeling, Experimentation, and the integrated ADT pilot are in development. The full Insights & Memory evidence graph is planned.

Specialist agents

Focused analytical workflows that form the foundation of the team vision.

04

Predictive Modeling

In Development

A specialist agent for building, validating, and explaining predictive models.