Why I started building
I started building BMail because recurring outgoing-email blocks were interrupting the team's communication. Fixing the transport was only the first step: people still needed a simple interface, clear permissions, and useful feedback. I turned that operational problem into a lightweight product that the team could use, with a foundation for a more complete business email service.
The challenge
Email reliability was an operational problem with a product gap behind it. Shared sending reputation, domain-authentication configuration, mailbox routing, and sender verification could each affect a message's journey. The team needed a clear way to compose and submit messages, use approved identities, and understand what had happened without managing SMTP credentials or interpreting server logs.
What I built in the first release
- Build the team experience: individual logins, administrator-managed access, approved sender identities, a composer with To/CC/BCC, attachments, and editable starter templates. Enforce a sender copy in CC on the backend so correspondence also reaches the user's existing mailbox.
- Design the application around explicit workflow state: a persistent sending queue, duplicate-submit protection, personal history, and an audit trail. Distinguish relay acceptance, partial rejection, failure, and uncertain completion so a connection interruption does not trigger an unsafe automatic resend.
- Keep the initial service lightweight: use Flask and server-rendered templates with one background worker, SQLCipher-encrypted SQLite storage, and server-side SMTP credentials. Add sender authorization, CSRF protection, password hashing, and sending limits around the workflow.
- Connect product and mail operations: integrate authenticated SMTP over verified STARTTLS, work through SPF/DKIM/DMARC alignment and same-domain routing issues, and preserve existing mailboxes for incoming messages and replies.
- Deliver the application as a Docker deployment with persistent data and HTTPS access through Cloudflare Tunnel. Migrate existing accounts and message history, then verify the browser experience and SMTP connectivity.
The system, at a glance
- 01Team login and approved sender selection
- 02Compose message and attachments
- 03Authorization and persistent queue
- 04Authenticated SMTP submission
- 05Submission status, history, and audit trail
What this work connects
A deployed internal service that gives the team an application for controlled business-email sending, rather than requiring direct access to mail-server configuration. The current setup uses infrastructure under the team's control and avoids an additional email-platform software subscription, while hosting and operations remain part of running the service.
What shaped the product
Infrastructure becomes a service when people can use it without understanding the servers underneath. Account management, sender permissions, composition, and history were product requirements, not finishing touches.
A successful SMTP submission is not the same as arrival in a recipient's inbox. Authentication, routing, sender verification, and later bounces are distinct concerns, so the interface must describe the status it actually knows.
The smallest useful release made the product practical to operate. A lightweight application and encrypted local storage supported the internal workflow without introducing a separate frontend build, external database, or queue service.
The full-service direction
Proposed next phases. The current release handles outgoing communication; receiving and the capabilities below are product directions, not features already delivered.
- Receiving and conversation workflows: explore inbound processing, reply threading, and shared-inbox capabilities to bring sending and receiving into one team workspace.
- Domain and team onboarding: develop guided domain verification, sender provisioning, and organization-level access boundaries for a reusable service beyond the initial team.
- Delivery feedback: integrate bounce and delivery-event processing, actionable diagnostics, and administrative reporting beyond the current submission-status view.
- Application integrations: add a developer-facing API, reusable templates, and event-driven transactional messaging for business applications.
- Service operations: extend retention controls, tested backup and recovery procedures, observability, and capacity planning as usage grows.
Selected contributions, not a claim of sole authorship. Client-specific configurations and operational data are intentionally excluded.