Reframe

I’m a product engineer at Reframe, working across customer-facing applications, internal tools, and backend services. I joined while much of this software was still early and became a primary contributor to the dashboard, onboarding, and AWS Lambda codebases.

My work spans about 2,000 authored commits across more than 20 independently deployed applications and a substantial serverless backend. I work with the CEO, coaches, support operators, clinical partners, and operations teams on product requirements, interface design, data models, implementation, and production support.

I rebuilt the support queue and ticket workspace around live customer data. The queue includes compound filters, saved views, assignment, presence, SLA tracking, bulk actions, and operator analytics. Tickets bring together messages, internal notes, attachments, reply drafts, account details, billing history, refunds, cancellations, and delivery status.

I built dedicated Lambda services for ticket access, customer context, audits, analytics, events, and notifications. Email handling includes scheduled Gmail ingestion, threaded replies, deduplication, delivery reconciliation, opt-outs, attachment checks, malware scanning, and resumable backfills.

The AI-assisted support pipeline handles taxonomy, duplicate detection, known-issue retrieval, and reply drafting. Confidence scoring, operator-curated examples, evaluation, and human review are included in the workflow.

Zacal is Reframe’s in-house scheduling platform. I developed event types, coach assignments, availability rules, holidays, intake questions, booking buffers, pricing, and role-specific administration.

The public booking flow supports invitations, cancellations, rescheduling, paid sessions, coach and customer time zones, and daylight saving changes. Atomic booking prevents two requests from taking the same slot.

Google Calendar, Apple Calendar, and Calendly integrations handle synchronization, webhooks, conflicts, and event write-back. I also built a workflow engine for booking-related messages, with templates, previews, test triggers, calendar attachments, and delivery logs. Recovery services cover calendar drift, expired bookings, interrupted requests, and reconnects.

I built and maintained onboarding products for Med+ programs, including naltrexone, HRT, GLP-1, and dermatology. These flows cover product information, intake, branching eligibility, account creation, consent, checkout, scheduling, and handoff to clinical systems.

A shared navigation and persistence model supports backtracking, route guards, deep links, returning users, cross-flow prefill, per-tab state, and webview-safe saves. I connected anonymous entry, Cognito accounts, submissions, and checkout, with attribution, flow-version tracking, and recovery for interrupted sessions.

Lambda services integrate DynamoDB, Hasura, Stripe, RevenueCat, Tellescope, and OpenLoop. The operations dashboard includes versioned funnels, page-flow graphs, individual path inspection, cross-system health checks, and record-level diagnostics. Backfills support previews before writes and report skipped, raced, and failed records.

I developed the Agora-based 1:1 meeting system, including authentication, secure entry, pre-join setup, device selection, camera and microphone controls, screen sharing, chat, network state, background effects, and picture-in-picture.

Meeting access connects directly to Zacal bookings. I also built major parts of the coach workspace: live conversations, unread state, receipts, reactions, attachments, scheduled and mass messages, saved drafts, notes, sessions, and calendar views. Permissioned client history is available within the same workspace.

I built a Snowflake-backed analytics product to replace Amplitude for internal use. It includes a declarative chart compiler, funnels, retention, cohorts, breakdowns, saved analyses, caching, user lookup, and an AI query assistant.

The migration synchronized a 1,352-chart analytics estate. A parity harness verifies chart definitions and SQL. Other analytics tools cover coach response times, workload, retention, meetings, users, subscriptions, and Stripe revenue, with filters, drill-downs, documentation, and exports.

I built Reframe Institute tools for courses, modules, learners, cohorts, mentors, assignments, peer review, and rubric grading. The media pipeline connects S3, MediaConvert, transcription, and subtitle editing.

Additional work includes content and challenge publishing, forum moderation, marketplace administration, surveys, partner programs, podcast publishing, and account creation. These products include validation, audit history, bulk operations, progress reporting, and recovery paths where needed.

I consolidated the dashboard suite into a pnpm monorepo with shared TypeScript UI, GraphQL, subscription, and data-access packages. Applications retain independent deployment boundaries. Path-aware GitHub Actions validate and deploy only the applications and shared packages affected by a change.

Dashboard Mod centralizes access management across the internal application fleet. It covers role and feature grants, bulk previews, revocation history, isolated super-admin access, authentication audits, and deployment visibility.

I standardized OAuth, server-verified sessions, Hasura roles, allowlisted operations, KMS-backed secrets, and authenticated Lambda calls. Test coverage includes unit, integration, browser, permission, parity, and failure-path checks. Health views, audit feeds, dry runs, repair actions, and controlled backfills support production maintenance.

This is ongoing employment. The interfaces and underlying health and customer data are proprietary, so this page describes the systems and my contribution without publishing internal screens or customer records.