Product workflow
Capture → clarify → plan → focus → adjust
Tools and stack
What it demonstrates
Problem and goal
Students often capture tasks, schedule work, and run focus sessions in separate tools. Stride explored one connected workflow where tasks could be clarified, scheduled into persisted focus blocks, completed in contextual focus sessions, and adjusted from the resulting task and focus context.
Project context
Stride was built as a full-stack portfolio MVP for students managing coursework, deadlines, projects, and study sessions. The repository does not establish a production user base, adoption, measured impact, or production-grade reliability.
Core product surfaces
Project outcome
Stride brings task capture, calendar planning, focus sessions, project organization, authentication, relational data, and realtime synchronization into one end-to-end MVP rather than presenting those capabilities as disconnected interface demos.
Architecture and data flow
Clerk protects non-public routes and supplies authenticated identity. Browser and server Supabase clients forward a Clerk access token, while Postgres row-level security remains the data-access boundary.
DataProvider loads profile preferences, accessible projects, and aggregate focus/task statistics.useTaskDataset loads route-level tasks, planning blocks, labels, attachments, memberships, and focus context.Key technical decisions
Keep capture, planning, and execution on shared data
Scheduled blocks and focus sessions retain work context instead of becoming isolated calendar or timer records.
Use RLS as the authorization boundary
Clerk supplies identity while Supabase policies distinguish user-owned records from project-member access.
Persist planning blocks
Planning is modeled as durable focus blocks rather than transient calendar interface state.
Reconcile optimistic updates with Realtime
Several flows combine local updates, database responses, subscriptions, and explicit echo suppression.
Keep AI bounded and optional
The server route authenticates, limits, validates, and sanitizes its context rather than exposing a general client-side prompt.
Preserve a final schema reference
The final SQL reflects the remote schema but is not a guaranteed one-command rebuild or complete external-service setup.
Project screens follow the core product story from fast capture and structured task detail through calendar planning, focus execution, and project organization.
Quick Add supports project, deadline, priority, estimate, reminder, recurrence, and label syntax, with optional attachments.
The detail surface handles the broader task model without forcing every field into the main task list.
Day, week, and month views place persisted planning blocks beside task context.
Focus and break modes persist completed sessions with available task, project, and planned-block context.
The project catalog summarizes workspaces before moving into project-level task organization.
Project workspaces support list and board organization, sections, task movement, and collaboration foundations.
Constraints and limitations
Stride was built as a portfolio MVP, not a production SaaS. The final version demonstrates a connected task, planning, and focus workflow, but it depended on external services such as Clerk, Supabase, OpenAI, Sentry, Vercel, and PostHog being configured correctly.
The app included foundations for collaboration, realtime updates, file storage, and AI-assisted daily summaries. I would present those as implemented technical foundations, not as production-proven systems.
The archived repository is best understood as evidence of full-stack product engineering, workflow design, and system integration. If I continued the project, I would narrow the scope, improve end-to-end test coverage, simplify deployment setup, and validate collaboration flows with real multi-user usage.
What I learned