Tools by task
Explore all toolsBuild a causal diagramDAG Studio Map study timingStudy Design Studio Plan a study & assess dataSPIFD2 Workspace Review reporting completenessTARGET Checklist Assess risk of biasROBINS-I Find tools & integrationsRWE MCP RegistryBuild your AI fluency
The Lab
Reading List
This Week's Picks Archive How It Works
Writing
About
Contact

Black Swan

Causal Labs

Software and methods
for better
real-world evidence.

We build tools for causal reasoning, study design, and evidence assessment. Domain expertise and AI, brought into workflows researchers can inspect and review.

Explore Tools Work With the Lab
Scroll
Built in the Lab

Start with the work
you need to do.

Study Design Studio

Make the timing
of your study visible.

Build assessment windows around an index event. Keep the reasoning with the design, then export an editable diagram for review.

Open the StudioExplore the workspace
Study Design Studio reference example showing assessment windows around an index eventStudy Design Studio · reference example
DAG Studio browser interface DAG Studio

Build a causal diagram

Draw a causal diagram and examine adjustment sets and backdoor paths. Work in the browser or connect an AI agent.

Explore tool v3.0
Study Design Diagram Studio with assessment windows around an index event Study Design Studio

Map study timing

Map eligibility, assessment, and follow-up windows. Export editable diagrams for protocols and presentations.

Explore tool WebMCP enabled
A feasibility heat map grid: study design elements scored one to five against candidate real-world data sources SPIFD2 Workspace

Plan a study & assess data

Carry a research question through study design and a structured assessment of candidate data sources.

Explore tool v1
A completed TARGET reporting checklist rendered by TARGET MCP, showing the title, the assessed manuscript and its citation, and a completeness bar TARGET Checklist

Review reporting completeness

Review a manuscript against the TARGET reporting checklist, with source-linked evidence for inspection.

Explore tool v0.1.2
A ROBINS-I V2 risk-of-bias assessment rendered by ROBINS-I MCP using synthetic data, showing the assessed result, the target trial specification, and the six domain judgements ROBINS-I

Assess risk of bias

Structure a ROBINS-I assessment with cited evidence and algorithmic judgements, ready for human review.

Explore tool v0.1.0
A hand-drawn diagram of how an MCP server works: a user prompts an LLM, which reaches external tools like web search, a database, and a code runner through an MCP server, then returns an answer RWE MCP Registry

Find tools & integrations

Find publicly available servers, skills, and connectors, with source-based notes on their relevance to RWE.

Explore tool Index 01 / 2026
The Lab

Scientific reasoning.
Made inspectable.

Black Swan Causal Labs is an independent lab building software and methods for real-world evidence. We start with the questions practitioners face: what to adjust for, how to design a study, and how to assess the evidence.

Our tools make assumptions, sources, and decisions visible. AI helps with the work; methodological judgement stays with the researcher.

Meet the founder

See the reasoning

Inspect the diagram, design choices, or cited evidence behind an output.

Explore evidence-linked assessment ↗

Understand the limits

Read each tool’s methods, validation scope, and limitations before relying on it.

Read the assessment approach ↗

Build on the work

Explore open-source projects, inspect implementations, and discuss a pilot with the lab.

Explore the source ↗
From the Lab’s Writing

Latest Thinking

The Delta: From AI-Naive to AI-Fluent in Pharmacoepidemiology
Crossing the AI Delta
New

The Delta: From AI-Naive
to AI-Fluent

Becoming AI-fluent does not mean mastering the stack. The next useful application in pharmacoepidemiology will probably come from a domain expert who knows which part of the job is worth fixing.

 Is Prompt Engineering Dead?
Prompting & Model Intent
Read

Is Prompt
Engineering Dead?

Four prompting strategies run on two LLM families, one HARPER protocol template task. The skill did not die. It moved from structuring the prompt to auditing what the model decided on your behalf.

 AI Agents and Idiot Sandwiches
Agent Architecture for RWE
Read

AI Agents and
Idiot Sandwiches

A note on agent architecture for RWE teams. Why one agent doing everything cooks like a stressed-out home cook, and what the brigade model offers instead.

 Every Study Is on the Causal Ladder
Causal Inference in Practice
Read

Every Study Is on
the Causal Ladder

Pearl's ladder gives us a way to read any paper with one question first: which rung is this research question on, and does the design have the equipment to answer it?
The weekly reading list

Stay close to the evidence.

Five papers on real-world evidence, causal inference, and AI, with a practitioner’s perspective.

Read this week’s picks
Work with the lab

Bring the work
into your organization.

Collaborate on a study, pilot a tool, or build your team’s AI capability.

Study design & evidence strategy

Causal inference, study architecture, and interpretation of RWE guidance.

AI systems & integration

Program design, agent workflows, and MCP development for evidence teams.

Training & enablement

Practitioner-led learning, workflow redesign, and support for AI adoption.

Advance Real World Evidence
Together.

Tell us what you’re working on: a tool pilot, a research collaboration, an advisory engagement, or training for your team.

Based In Los Angeles, California
Focus Pharmacoepidemiology & Real-World Evidence
Entity Black Swan Causal Labs, LLC