kayla’s scrapbooklet’s talk ↗
03 / a project pulled from the pocket
Chat → CRM

From conversations to a useful CRM.

A CRM is only as useful as what makes it into the record. This workflow starts in WhatsApp, finds relevant business context, and brings proposed updates back to a person before they are committed.

focus

Workflow design · backend · operational debugging

built with

AI workflows · CRM integration · Validation · Monitoring

the whole journey, at a glance.

a little map of the productfollow the thread ↘

Business context starts in the chat people already use.

Interactive product-flow illustration · tap a step to explore

01 / the starting point

What needed to work.

Important details lived inside everyday chats. Asking busy agents to repeat all of that work in another interface created a gap between what was happening and what the CRM knew.

02 / my part

The work behind it.

I contributed to the chat-to-CRM workflow, including staged findings, confirmation paths and operational debugging for incomplete or failed processing.

03 / an intentional choice

The decision that mattered.

Separate findings from writes. Group proposed changes into reviewable bundles and require confirmation. Monitor actual processing outcomes as well as scheduled jobs, and retain a retry path for work that did not finish.

a note I’m taking with me

An agent workflow is a product with states, ownership and recovery. Its success is not just whether a model can produce a plausible answer.

04 / project context

A little context.

Sanitized reconstruction of an internship workflow. The visual is illustrative and contains no real customer conversations or records. It shows the intended review flow, not a claim that incidents never occurred.