Real-world, working examples

Problem

• The leaders' standard work had not been mapped or updated.
• Gemba walks, shift handovers and issues were managed outside any common system.
• Supervisors and managers had no real-time information about the operation.
• Related documentation —Gemba Wall, daily warehouse inspection procedure— was out of date.

Solution

• Leaders' Standard Work: mapping and updating the standard work.
• Gemba Walk: daily walks across 3 shifts, digitized in the app.
• Shift handover log, built into the same app.
• Management of issues raised during the Gemba walk, with follow-up.
• AI agent in Microsoft Copilot built on corporate procedures, work instructions, manuals and the knowledge of leaders, supervisors, managers and experts, to support planning, organizing and deciding during the shift.
• Documentation update: creation and update of the Gemba Wall and of the daily warehouse inspection procedure, among others.

Result

• Project in progress, partially in production.
• ~20 leaders in Cartagena as direct users, across three warehouses.
• Gemba walks digitized across the 3 daily shifts.
• Operational information available in real time to supervisors and managers.

Problem

• The checklist was printed from an Excel template and filled in by hand in the field.
• The data was then transcribed into a computer.
• The report was consolidated and shared in a later step.
• The full cycle —inspection, transcription, report, sharing— could take a week.

Solution

• Real-time field app with a score for each item: compliant (2), partially compliant (1), non-compliant (0).
• Automatic calculation of the compliance percentage over the maximum possible score.
• Report generated immediately when the audit is closed.
• Automatic email notifications to leaders, supervisors and owners.
• Integration with the corporate action-plan management system, to follow up on findings.

Result

• Full cycle reduced from a week to a day.
• Printing the template and manual transcription eliminated.
• Report available immediately when the inspection closes.
• 5–6 auditors working in the field on the app; ~10 direct recipients of the report.
• 2 audits per month sustained.

Problem

• There were systems to assign loads and orders, but loading was documented on paper.
• Vehicle inspection forms were filled in by hand.
• Photos were taken on a phone and had to be downloaded, matched to the load record, organized and uploaded to a network folder.
• When a customer filed a claim, someone had to write manual reports, search for paper documents and gather photos.

Solution

• Checks and inspections filled in inside the app, at the moment of loading.
• Photo evidence captured from the app itself: ~20 photos per load on the loading lines, ~25 including the prior requirements.
• Real-time sync with SharePoint, with no manual download or filing.
• Consolidated report generated automatically when each load is finished.
• Digital load record available without anyone having to assemble it.

Result

• Complete digital record per load, available immediately.
• Paper forms, manual photo downloads and network-folder filing eliminated.
• Customer claims handled by looking up the record directly, with no ad hoc reports.
• Loads documented daily, with ~20–25 photos each.

Problem

• Commitments were written down by hand during the meeting.
• Consolidation happened afterwards, in Excel.
• Sending them to the owners was manual, by email.
• There was no central record and no tracking of each commitment's real status.
• Preparing the Visual Room (consolidation, presentations and results) depended entirely on one person.

Solution

• Commitments recorded in real time during the meeting, from the app.
• Automatic email notifications to the owner of each commitment.
• Automatic weekly delivery of the consolidated list of pending commitments.
• Visual Room automation module: information consolidation, presentation generation and results publishing with no manual intervention.
• Later scope expansion to manage any commitment in the department, not only those originating in Visual Room.

Result

• Central record of more than 100 commitments, with traceability and status tracking.
• Manual consolidation in Excel and manual emailing eliminated.
• Administrative load lifted from the person responsible for preparing the Visual Room.
• ~20 direct users and their teams working from the same source of information.

Problem

• Each warehouse handled its indicators separately, with no standard definition.
• Extraction and calculation were manual.
• The staff in charge spent roughly a week extracting and calculating each indicator.
• There was no single view of the department's performance.

Solution

• Structured data capture in SharePoint / Microsoft Lists forms and lists.
• Standard definition of the indicators, shared by every warehouse and area.
• Automatic calculation with Power Automate and Power Platform, with no manual extraction.
• Power BI dashboard with the department's indicators, updated in real time.
• Scope extended from the initial area to the whole department, including external warehouses.

Result

• Centralized, standardized indicators for 3 own warehouses, ~6 external warehouses and the department's areas.
• Manual calculation that took roughly a week per cycle eliminated.
• Automatic calculation and real-time visualization as the basis for decisions.
• More than 20 users on the system.

Problem

• The business was growing with no formal management structure.
• Inventory, sales and activities were recorded manually and scattered.
• Searching for information and recording it took time away from the operation.
• There was no budget for paid management software.

Solution

• Inventory management.
• Accounting.
• Digital product catalog.
• Task management.
• Social media content planning.
• Activity tracking and operational scheduling.
• Automations, reminders and alerts within the same workspace.
• Everything built on Notion, a free tool accessible to the client.

Result

• System implemented, paid for by the client and in use today.
• Better sales traceability and management than the previous manual records.
• Hours saved by reducing manual recording, searching and handling of information.
• Operation centralized in a single workspace, with no license cost for the client.

Problem

• A professional portfolio on a template doesn't prove design or development skills.
• The site had to be my own, consistent with a defined design system and maintainable over time.

Solution

• Site built with Next.js and TypeScript.
• Styling on Tailwind CSS, with custom design tokens (color, typography, spacing, radii, shadows, motion, breakpoints).
• Component library based on shadcn/ui, extended with the system's own components.
• Two brand themes derived from the same tokens.
• Site published to production.

Result

• Site built and published: https://cristian-castilla.com.
• Custom design system and design tokens, applied consistently across the product.
• Shows the full path from design system to published product.