MB

Selected work — developer documentation, shipped.

An annotated list of public docs and content systems I’ve led or built. Most are linked; the systems work (test frameworks, generation pipelines, automated cross-linking) is internal but described here.

Braintrust Data

June 2026 – Present

  • Query API reference — elevating /btql to a top-level API

    ReferenceAuthor

    New API reference for the /btql endpoint, which runs SQL queries against logs, experiments, and datasets. Covers request and response schema, pagination, timeouts, and runnable curl, TypeScript, and Python examples, each tested against the live endpoint.

    braintrust.dev

  • SQL charts — docs for a Public Preview launch

    InstructionalAuthor

    Wrote the docs for SQL charts from scratch: dashboard charts defined by a query, built with a visual builder or as SQL. Covers chart types and options, table charts, query requirements with example queries, charting results from the Query page, and building charts with Loop.

    braintrust.dev

  • Optimize costs — guidance for a customer email campaign

    InstructionalAuthor

    New guide that Marketing commissioned for an email campaign, scoped with a PM, Solutions Engineering, and DevRel. It shows customers how to cut both their Braintrust bill and their LLM spend: less duplicate logging, sampled online scoring, cheaper models and judges, batch evals, and how to measure the savings.

    braintrust.dev

  • Prompts — two monolithic pages into one navigable section

    InstructionalAuthor + information architect

    The largest restructure I’ve done at Braintrust. Prompts were documented on two long pages, one under Evaluate and one under Deploy, that had drifted apart and filled up with duplicate and overlapping content. I merged them into a single section of scannable sub-pages organized around how a prompt moves from authoring to production (create, template, use in code, version, manage), rewriting and retitling along the way and cutting the content about 20%.

    braintrust.dev

  • Access control — permission model restructure

    ExplanatoryAuthor + information architect

    Rebuilt the access control section and all its sub-pages around the permission hierarchy — organization, project, and object scope — plus built-in groups and a full permissions reference.

    braintrust.dev

  • Human review — score configuration and reviewer workflows

    InstructionalAuthor + information architect

    Restructured the human review section and sub-pages: configuring review scores, scoring traces and datasets, managing review work across a team, and combining scores from multiple reviewers.

    braintrust.dev

  • Experiments — SDK, UI, and CI/CD evaluation paths

    InstructionalAuthor + information architect

    Reorganized the experiments section around the three ways teams actually run evaluations — in code with the SDK, in the UI, and in CI/CD to catch regressions — plus analyzing and comparing results.

    braintrust.dev

  • SQL reference — SQL and BTQL query syntax

    ReferenceAuthor + information architect

    Restructured the SQL reference and sub-pages covering both SQL and the legacy BTQL syntax, the four execution surfaces (sandbox, Loop, bt CLI, API), and documented limitations.

    braintrust.dev

  • Docs evals for agents — Brainstore SQL

    SystemsBuilder

    Part of an ongoing team project that dogfoods Braintrust to test the docs. Using a Datadog MCP server, I mined production logs for the customer SQL questions that most often ended in Brainstore errors, and built a dataset of those questions with expected answers. Running evals on it in Braintrust produces scored experiments that show how well agents answer technical questions from the docs, and the failure patterns point to what the docs should fix next.

  • Docs infrastructure and automation

    SystemsBuilder

    Mintlify and AI-platform infrastructure the docs team runs on: agent skills, GitHub Actions workflows, shared Claude/agent settings, style-guide conventions, and automated gap analysis runs that flag coverage holes and drive what gets written next.

Snowflake

May 2023 – April 2026

  • Snowflake Python APIs — Developer Guide

    InstructionalAuthor, lead writer through Private Preview → Public Preview → GA

    End-to-end developer guide for Snowflake’s native Python APIs package: installation, connecting, core concepts, and task-based workflows.

    docs.snowflake.com

  • Snowflake Python APIs — API Reference

    ReferenceAuthor + tooling

    Code-generated API reference, regenerated from source on every API version release via a Python + Sphinx pipeline I built, which also ships the release notes.

    docs.snowflake.com

  • Snowflake Python APIs — Getting Started Tutorials

    InstructionalAuthor

    Getting-started tutorial series taking developers from a clean environment to working code against the Python APIs.

    docs.snowflake.com

  • Snowflake Python API Reference — automated SQL reference cross-linking

    SystemsDesigner + cross-functional lead (Doc Platform Engineering, Product Engineering, UX Design)

    Automated SQL reference cross-linking for the Python API Reference, reducing manual lookup effort and ensuring consistency across the API reference.

    Cross-link layer with SQL refs increased API reference page views ~30%.

    docs.snowflake.com

  • Snowflake Developer landing page rearchitect

    ExplanatoryCross-functional lead (Product, Engineering, Marketing, Legal)

    Rearchitected the Developer landing page from feature-listed to task-based, user-story-driven information architecture.

    Received positive feedback directly from customers through UX surveys and usability studies.

    docs.snowflake.com

  • Automated code-sample test framework

    SystemsArchitect (in partnership with product engineers)

    In-house CI/CD framework that executes every in-doc code sample against the current build before it ships, keeping reference content continuously verified against the live API.

  • Behavior Change Release (BCR) announcements

    ReferenceRelease manager / lifecycle owner

    End-to-end deployment of breaking-change announcements across the entire Snowflake documentation surface, coordinated with Product, Engineering, and stakeholder teams.

    docs.snowflake.com

Amazon Web Services

Oct 2018 – May 2023

  • Amazon QLDB — Developer Guide

    InstructionalAuthor (Private Beta → Public Preview → GA)

    Developer guide with code examples in Java, .NET (C#), Go, Node.js (JavaScript and TypeScript), and Python.

    ~21,000 monthly page visits, ~18,000 unique visitors.

    docs.aws.amazon.com

  • Amazon QLDB — API References

    ReferenceAuthor

    API reference content across all supported AWS SDKs, the AWS CLI, and AWS CloudFormation.

    docs.aws.amazon.com

  • Amazon QLDB — Console UI text, help panels, error messages

    In-productAuthor

    In-product copy across the AWS Management Console UI, help panels, and service error responses.

  • AWS re:Invent 2019 — QLDB workshop curriculum

    InstructionalCurriculum designer + co-instructor

    Designed hands-on code lab curriculum and co-taught QLDB workshops alongside AWS Solutions Architects.

    CSAT 4.2/5 across sessions.

  • Amazon Managed Blockchain — public docs

    ReferenceManager, Technical Documentation (2021–2023)

    Owned the public technical content for all Amazon Managed Blockchain services. Recruited, onboarded, coached, and edited the writers producing it.

    docs.aws.amazon.com

  • Amazon FinSpace — public docs

    ReferenceManager, Technical Documentation (2021–2023)

    Owned the public technical content for Amazon FinSpace. Recruited, onboarded, coached, and edited the writers producing it.

    docs.aws.amazon.com