// SELECTED WORK

Built around real operational problems.

Systems and hackathon builds that show how I work: understand the workflow, translate it into a product, connect the moving parts, and keep the result useful to the people operating it.

PROOF, NOT A TOOL LIST

CLIENT-SAFE DETAILS · PROPRIETARY WORK RESPECTED
PUBLIC LAB · EN / ES

From a business problem to a clear next step.

Try an AI discovery demo: three questions, an opportunity brief, and the unknowns still worth validating.

Open demo

// HACKATHON BUILDS

AI for everyday problems.

Recycling, internal tools, and digital commerce. Three projects connecting a specific user need with a product experience.

01 / CIRCULAR ECONOMY

UTEL + Google AI hackathon

AI-assisted recycling in Mexico

HACKATHON RESULT
Semifinalist

An AI-based app connecting people who recycle with recycling centers in Mexico. The concept addressed two everyday concerns: finding fairer recycling options and understanding where recyclable materials should go.

  • Connections with recycling centers
  • Fairer access to recycling options
  • Appropriate destinations for materials
My role & project goal

MY CONTRIBUTION

Built an AI-based application focused on connecting people who recycle with recycling centers in Mexico, with a simpler experience at the center of the project.

INTENDED IMPACT

Make recycling easier and fairer while helping people find appropriate destinations for their materials, with the aim of keeping recyclables out of general waste.

02 / CLOUD + AI

AWS hackathon · Around 2024

Cloud and AI tools for business

HACKATHON RESULT
Quarterfinalist

A team-built suite of cloud and AI-based products for a company, ranging from VPNs to internal tools. The project explored alternatives to expensive subscriptions through tools shaped around the company’s needs.

  • VPNs and internal tools
  • Cloud and AI-based products
  • Alternatives to costly subscriptions
My role & project goal

MY CONTRIBUTION

Worked with a team to create a suite of cloud and AI-based products, including VPNs and internal tools, during an AWS hackathon.

INTENDED IMPACT

Explore how purpose-built cloud and AI tools could replace expensive subscriptions and give a company software better aligned with its internal workflows.

03 / COMMERCE + AI

E-commerce hackathon for a Mexican brand

Virtual try-on from a product QR code

HACKATHON RESULT
Approx. US$2,000 in scholarships shared among 5 team members

A team-built shopping concept for a Mexican e-commerce brand. Product-specific QR codes opened an AI virtual try-on experience using a phone camera, with size recommendations informed by garment measurements.

  • Product-specific QR entry points
  • Phone-camera virtual try-on
  • Garment-based size recommendations
My role & project goal

MY CONTRIBUTION

Co-built a product experience with a team, combining QR entry points, AI-assisted virtual try-on through a phone camera, and recommendations based on garment measurements.

INTENDED IMPACT

Help shoppers make more informed size choices before purchasing, with the aim of reducing fit-related returns and making online clothing shopping easier.

Project goals describe the intended benefits, not measured results from a production deployment.

01 / CASE STUDYOFFLINE-FIRST · PROPRIETARY CONSULTING TOOL

ISO 27001 Gap Assessment Platform

MY ROLEProduct design and implementation

WHAT IT MADE POSSIBLE

Delivered a portable assessment application that replaced SecurePath’s Excel workflow and kept engagement data on the consultant’s computer.

Annex A controls
93
Self-contained HTML file
1
Server calls
0

THE STARTING POINT

Consultants carried out the assessment in spreadsheets. The next tool needed to bring controls, evidence, observations, and reporting into one workflow without requiring an external service.

WITH THE SYSTEM

A consultant can open one HTML file, review the controls, record evidence and maturity scores, and prepare assessment outputs locally, including an executive report and Statement of Applicability support.

HOW IT WORKS

Walk through the workflow.

ILLUSTRATED WORKFLOW

Explore the system’s inputs, decisions, and outputs.

STEP 01 / 03

Turn a framework into a guided review.

I organized the assessment around the 93 Annex A controls so consultants could review the framework inside the application while keeping engagement data local.

INPUT

The organization’s current practices and assessment scope

OUTPUT

A structured review of the applicable controls

Select a step to see what happens.
Technical decisions & scope

// PROBLEM

SecurePath was running its assessment workflow in spreadsheets. The firm needed a portable assessment tool that consultants could open locally and use without sending sensitive engagement data to an external service.

// MY CONTRIBUTION

Built the entire platform as a self-contained HTML file. The application intentionally makes no server calls and requires no internet connection, so an authorized consultant can take the file directly into an engagement and keep the workflow local.

// DELIVERED CAPABILITIES

  • One portable, self-contained HTML application
  • No internet connection, backend, or server callback required
  • 93 Annex A controls in one guided workflow
  • Evidence tracking, maturity scoring, observations, and prioritized gaps
  • Executive reporting and Statement of Applicability support
  • Replaced the firm’s previous Excel workflow

// SYSTEM

Self-contained HTMLLocal-only processingOffline operationPortable deployment

The original application and its assessment content are proprietary to the consulting firm. The public demo uses fictional data and a newly built interface; the original source code remains private.

02 / CASE STUDYFOUNDER BUILD · RETAIL AUTOMATION

Device Trainer Sales & Delivery System

MY ROLECommercial model, automation, and backend engineering

WHAT IT MADE POSSIBLE

Connected retailer sales activity to automated coupon delivery, bringing the commercial model and the product’s fulfillment workflow together.

Retail sale detection
POS
Coupon delivery trigger
Automatic
Retail partner sales model
Channel

THE STARTING POINT

Selling a phone training membership through wireless retailers required a way to recognize the right transaction and initiate customer delivery without adding a manual handoff at every store.

WITH THE SYSTEM

The connected workflow detects a Device Trainer SKU sale through the retailer POS integration and triggers coupon delivery, using the IQmetrix API, n8n, Make.com, and Supabase.

HOW IT WORKS

Walk through the workflow.

ILLUSTRATED WORKFLOW

Explore the system’s inputs, decisions, and outputs.

STEP 01 / 03

Start with the sale that already happens.

I connected the product workflow to retailer transaction activity through the IQmetrix API so a Device Trainer SKU sale could initiate fulfillment.

INPUT

Retailer POS activity containing a Device Trainer SKU sale

OUTPUT

A recognized sale that starts the delivery workflow

Select a step to see what happens.
Technical decisions & scope

// PROBLEM

A membership-based phone training product sold through wireless retailers needs channel partners, transaction awareness, and dependable fulfillment without adding manual work at every store.

// MY CONTRIBUTION

Designed the commercial model and built the connected automation behind the product line, linking retailer activity to fulfillment while continuing backend product work.

// DELIVERED CAPABILITIES

  • Detects Device Trainer SKU sales in the retailer POS workflow
  • Triggers coupon delivery automatically
  • Connects n8n, Make.com, Supabase, and the IQmetrix API
  • Supports a channel-sales model through wireless retailers

// SYSTEM

IQmetrix APIn8nMake.comSupabase

This system connects product engineering directly to channel revenue and customer delivery.

03 / CASE STUDYAI PRODUCT · PROPRIETARY WHOP APP

NEPQ Practice Simulator

MY ROLEAI training product design and implementation

WHAT IT MADE POSSIBLE

Delivered an adaptive AI customer for independent sales practice inside a paid Whop subscription experience.

Conversational roleplay
AI customer
Responses and objections
Adaptive
Subscription delivery
Whop

THE STARTING POINT

A salesperson’s roleplay practice depended on another person being available. A fixed script could not respond to the seller’s choices or track the required discovery points through a conversation.

WITH THE SYSTEM

The seller can hold a two-way conversation with an AI customer. The scenario adapts to the seller’s input and continues until the required NEPQ discovery points are covered.

HOW IT WORKS

Walk through the workflow.

ILLUSTRATED WORKFLOW

Explore the system’s inputs, decisions, and outputs.

STEP 01 / 03

Make a practice partner available on demand.

I built the experience around an AI that plays the customer, allowing a salesperson to rehearse a discovery conversation without waiting for a live partner.

INPUT

A practice scenario and the seller’s opening conversation

OUTPUT

A two-way roleplay with an AI customer

Select a step to see what happens.
Technical decisions & scope

// PROBLEM

Salespeople need realistic roleplay practice even when a manager, trainer, or teammate is not available. Static scripts cannot adapt to what the seller says or verify that the full discovery process was covered.

// MY CONTRIBUTION

Built an AI-driven training experience inside a Whop subscription product. The AI acts as the customer, holds a two-way conversation, introduces new responses and objections, and continues the scenario until the seller reaches the required discovery points.

// DELIVERED CAPABILITIES

  • Conversational AI customer roleplay
  • Adaptive responses based on the seller’s input
  • Tracks required NEPQ discovery points through the scenario
  • Allows independent practice without a live roleplay partner
  • Delivered inside a paid subscription experience on Whop

// SYSTEM

Conversational AIPrompt systemsAdaptive roleplayWhop subscription app

This application is proprietary to the subscription service, so the live experience and source cannot be shared publicly.

Need someone who can sit with the user, understand the workflow, and still ship the system?

Let’s talk
AXEL VEGA MÉNDEZ · 2026AI SOLUTIONS · AUTOMATION · SECURE DELIVERY