Skip to content
Linfy
Case studies

Linfy Beacon

Local prototype

A clearer way to find businesses that may be hard to find online.

Linfy Beacon is a locally developed prototype with separate modules for public business discovery, website assessment, and AI-assisted outreach drafting. Integrated orchestration remains planned.

Have a question before starting? Message us directly on WhatsApp.

Conceptual artwork showing Linfy Beacon discovery, assessment, drafting, and record-keeping modules
Conceptual case study artwork. It illustrates the intended workflow, not a live product screen. View full-size illustration (opens in a new tab)

The problem

Some small businesses are harder to discover than they need to be.

A business without a current website or clear online presence can be difficult for a prospective customer to find. Identifying that gap manually across local areas is slow, inconsistent, and difficult to track responsibly.

Beacon was designed as a practical exploration of how public business data, website signals, structured records, and careful drafting could support a more focused local outreach workflow.

Discovery, with context

A business-discovery workflow built for review.

The local discovery module collects public business information for separately invoked assessment and outreach modules.

Conceptual Linfy Beacon business discovery dashboard showing generic categories and review states
Conceptual interface. Generic labels illustrate a review queue and do not represent real businesses or contact records. View full-size illustration (opens in a new tab)

Business discovery

A locally run Python module queries public map data, inserts new lead records, and skips matching name-and-address conflicts. Records without an address may still duplicate.

Website assessment

A separate assessment module generates notes from available website data or a missing website URL in the source record.

Considered drafting

A drafting module uses a stored template and an AI-assisted rewrite of assessment notes to prepare an email draft.

Outbound groundwork

A local sender module attempts to log sends on a best-effort basis. A send may succeed even if its log entry fails.

Assessment, not assumption

Review web-presence signals before drafting a message.

The assessment process generates notes from available website data or a missing website URL in the source record. A missing URL does not prove that a business has no website, and automated findings need verification before they are used in outreach.

The prototype does not enforce human review. Review is recommended before outreach, alongside appropriate production safeguards.

Conceptual Linfy Beacon website assessment interface with generic signals and review note
Conceptual assessment interface. It demonstrates an illustrative review process, not a production assessment result. View full-size illustration (opens in a new tab)

The workflow

Separate local modules, with orchestration still planned.

Discovery, assessment, drafting, and sending exist as separate local modules. Integrated orchestration and event handling remain planned; send logging is best-effort and can fail even when a send succeeds.

Conceptual diagram of separate Linfy Beacon modules and planned orchestration
Conceptual workflow. The connections represent planned orchestration; reply routing and event handling are not presented as live capabilities. View full-size illustration (opens in a new tab)
Conceptual Linfy Beacon AI-assisted email drafting interface without real recipient data
Conceptual drafting interface. No real recipient, message, business record, or delivery outcome is shown. View full-size illustration (opens in a new tab)

AI-assisted drafting

Turn assessment notes into an outreach draft.

Beacon uses a stored template and prompts an AI model to rewrite assessment notes for an outreach draft. Notes can contain multiple findings, and the generated rewrite is not automatically validated.

  • The model is prompted to produce a concise rewrite; the result is not guaranteed.
  • Template placeholders keep the drafting structure consistent.
  • Human review is recommended but not enforced by the prototype.

Technology stack

Focused tools for a practical prototype.

The prototype uses services chosen for separate local modules. The stack below describes the local implementation, not a public SaaS platform.

Python

Local discovery, assessment, drafting, and sender modules.

Supabase

Lead state, templates, and outreach-log data model.

OpenStreetMap

Public business discovery by category and mapped area.

Tavily

Website-content extraction used by the assessment module.

Anthropic

Prompted rewrite of assessment notes for an email draft.

Resend

A local email delivery integration module.

Development journey

Built in stages, with the next boundaries visible.

Beacon is being developed as a careful local prototype. The next work is about validation, operating controls, and evidence rather than presenting unfinished automation as complete.

  1. 01

    Foundation

    A local data model and separate discovery, assessment, drafting, and sender modules establish the prototype's components.

  2. 02

    Current focus

    Review assessment quality and complete controlled sender verification. Outreach logging is best-effort rather than guaranteed.

  3. 03

    Future roadmap

    Add integrated orchestration, scheduling, stronger send safeguards, unsubscribe handling, and event processing before representing those capabilities as active.

Start with the workflow

Could a more considered system make your work easier to run?

An Automation Audit is Linfy's current starting point for understanding one workflow before deciding whether a technical next step is appropriate.