Production AI Portfolio

AI Engineer building production intelligence systems.

I design ML pipelines, LLM systems, and data products that turn behavior, geospatial, and operational data into measurable business decisions — from loyalty intelligence and brand health monitoring to enterprise GenAI platforms.

Current role

At Buzzebees, I build AI systems for loyalty and customer engagement platforms, focusing on brand health intelligence, customer segmentation, reward optimization, churn signals, and marketing automation analytics.

View WorkDownload ResumeContact / Book Session

Answer First

Kane builds production AI systems for loyalty, geospatial, and enterprise operations.

Watcharapon “Kane” Weeraborirak is an AI Engineer in Bangkok building ML pipelines, LLM platforms, and data products for business decisioning.

His current Buzzebees work focuses on loyalty intelligence: brand health, segmentation, reward optimization, churn signals, and marketing analytics.

His product-builder work includes Goatie, a LINE-first AI accounting assistant for Thai SMEs built with accountant-led business insight.

Impact Snapshot

The systems I build convert noisy data into operational decisions.

Concise focus areas across production AI, enterprise platforms, geospatial intelligence, automation, and cloud architecture.

Focus 01

Production ML Pipelines

Batch and monitored inference systems that convert raw customer, brand, and operational signals into decision-ready outputs.

Focus 02

Enterprise GenAI Platform

Secure internal LLM access patterns with identity, retrieval, governance, prompt operations, and evaluation loops.

Focus 03

Loyalty Intelligence

Customer segmentation, reward optimization, churn/retention signals, and campaign analytics for loyalty platforms.

Focus 04

Geospatial AI Systems

Satellite, drone, raster, and vector pipelines for environmental intelligence, carbon readiness, and spatial operations.

Focus 05

Workflow Automation

Event-driven automations that move model outputs into alerts, summaries, CRM workflows, and marketing actions.

Focus 06

Cloud Architecture

Observable AWS and Azure platforms with infrastructure patterns for reliable data, ML, and product delivery.

Current Work at Buzzebees

Loyalty intelligence systems that connect behavior signals to marketing action.

I build production AI systems for loyalty and customer engagement platforms, with a focus on brand health intelligence, segmentation, reward optimization, churn and retention signals, and marketing automation analytics.

Accessible Pipeline

  1. 01Data Sources
  2. 02Feature Engineering
  3. 03Weak Labels / ML Model
  4. 04Batch Inference
  5. 05Business Summary
  6. 06Marketing ActionAction

Brand Health Intelligence from campaign, customer, merchant, reward, and engagement signals.

Customer segmentation and churn/retention indicators that help teams prioritize lifecycle actions.

Reward optimization analysis that links redemption behavior with business and engagement context.

Azure ML, MLflow, Databricks, Azure Blob, batch inference, and model monitoring patterns.

What I Do

Systems that stay fast, observable, and elegant.

Four pillars that guide every build: grounded discovery, reproducible delivery, and relentlessly measured outcomes.

LLM and vision models tuned for production.

AI Engineering

Design and deploy latency-sensitive inference pipelines with evaluation harnesses, guardrails, and cost observability.

  • LLM orchestration (Bedrock, OpenAI, LangChain, MCP)
  • Retrieval pipelines with vector + PostGIS hybrid search
  • Model evaluation loops and rollout automation

Geospatial ETL that scales from field sensors to dashboards.

Data Engineering

Stream, transform, and warehouse multi-modal data with schema governance and lineage built-in.

  • Serverless ETL / ELT orchestrations with AWS Step Functions
  • PostGIS optimization for spatial joins and raster analytics
  • S3, Glue, Athena, and Lake Formation governance

Close the loop between insights and action.

Workflow Automation

Operationalise ML outcomes using event-driven jobs, notifications, and low-code automations.

  • MLflow model lifecycle with approvals and stage transitions
  • n8n and custom workers for alerts, emails, and syncs
  • Observability via OpenTelemetry, CloudWatch, and Grafana

Secure, repeatable infrastructure as code.

Cloud Architecture

Architect AWS serverless platforms with CI/CD, multi-account guardrails, and zero-downtime deploys.

  • AWS CDK / CloudFormation blueprints
  • Security best practices (CIS, Well Architected)
  • Edge delivery with CloudFront, global networking, WAF
Selected Case Studies

Business-facing AI systems, not demo artifacts.

Each case study appears like a product module in the same intelligence world: problem, role, system design, AI approach, architecture, impact, stack, and lessons learned.

Current Work · Loyalty Intelligence

Buzzebees Brand Health Intelligence

Production ML pipeline patterns for turning loyalty, campaign, reward, and customer engagement signals into brand health summaries and marketing actions.

  • Azure ML
  • MLflow
  • Databricks
  • Azure Blob
  • Batch Inference
  • Model Monitoring
  • Frames brand health as actionable business signals instead of disconnected campaign reports.
  • Connects feature engineering, weak labels, model outputs, and summaries into repeatable decision workflows.
  • Supports marketing teams with customer segments, churn/retention indicators, and reward optimization context.
Read case study
Product Builder · LINE-first Accounting Workspace

Goatie — AI Accounting Assistant for Thai SMEs

A LINE-first accounting assistant that helps small business owners issue documents, scan receipts, organize expense records, and understand income and expenses from the tools they already use every day.

  • AI Product
  • SME Accounting
  • LINE Workflow
  • Document AI
  • Receipt Extraction
  • Payment Voucher
  • Turns LINE into a lightweight accounting workspace for Thai SME workflows.
  • Helps owners issue documents, scan receipts, organize expense evidence, and understand business summaries faster.
  • Combines accountant-led business insight with AI product engineering and mobile-first execution.
Read case study
Past Role · Enterprise GenAI Platform

AI4ALL @ Thaicom

Secure enterprise GenAI foundation integrating chat interfaces, identity, retrieval, governance, and model access patterns.

  • LibreChat
  • AWS Bedrock
  • RAG
  • SageMaker
  • Entra ID
  • LangChain
  • Created a governed entry point for enterprise GenAI experimentation.
  • Connected identity, knowledge retrieval, model access, and prompt operations into one platform pattern.
  • Helped teams evaluate LLM workflows with clearer security and operational boundaries.
Read case study
Geospatial AI for Forest Integrity

Carbon Watch

Satellite, drone, and ground-truth data pipelines for carbon readiness, forest health analysis, and environmental reporting.

  • AWS Lambda
  • SageMaker
  • PostGIS
  • Raster Analytics
  • S3
  • CloudWatch
  • Unified geospatial data sources into a more inspectable analytics workflow.
  • Supported carbon readiness and forest health analysis with repeatable data processing patterns.
  • Connected spatial data engineering with ML inference and executive reporting needs.
Read case study
Serverless Hospitality Platform

EasyStay Asia

Booking, payment, availability, messaging, and operations automation for a hospitality product.

  • Next.js
  • AWS Lambda
  • DynamoDB
  • OPN Payments
  • n8n
  • CDK
  • Turned booking and operations workflows into a connected digital product.
  • Reduced manual coordination through payment, messaging, and reporting automations.
  • Created a platform foundation that can evolve with hospitality business needs.
Read case study
Learning Platform Automation

Pasa Education

Course management, learner progress, content delivery, billing, and lifecycle messaging for an education platform.

  • Nuxt
  • Supabase
  • Serverless
  • n8n
  • Vercel
  • Created a structured course and learner management workflow.
  • Automated lifecycle communication across common student touchpoints.
  • Gave operators clearer visibility into course engagement and student progress.
Read case study
Journey

A narrative from automation to production intelligence.

Roles and environments that shaped how I build: enterprise chatbots, geospatial AI, independent product builds, and current loyalty intelligence systems.

Current

AI Engineer · Current

BUZZEBEES

Building AI systems for loyalty and customer engagement platforms, focused on brand health intelligence, segmentation, reward optimization, churn signals, and marketing automation analytics.

  • Designing Brand Health Intelligence workflows that convert loyalty and campaign data into business summaries.
  • Building ML pipeline patterns across feature engineering, weak labels, batch inference, and monitoring.
  • Working with Azure ML, MLflow, Databricks, Azure Blob, and production analytics patterns.
Azure MLMLflowDatabricksAzure BlobPythonAnalytics

Jul 2022 – Past role

AI Engineer · Jul 2022 – Past role

THAICOM PLC

Built geospatial data engineering and LLM deployment patterns for environmental intelligence, Carbon Watch, disaster prediction initiatives, and enterprise GenAI enablement.

  • Engineered async geospatial workflows across Lambda, SageMaker, and S3 powering Carbon Watch and climate risk analytics.
  • Built serverless ELT with AWS Glue + PostGIS for automated spatial ingestion, validation, and warehousing.
  • Integrated secure GenAI access with Entra ID, Bedrock, Anthropic, and OpenAI via LibreChat, OpenWebUI, MCP, and LangChain.
AWSSageMakerBedrockPostGISLangChainCloudFormation/CDK

Jul 2020 – Jul 2022

AI Developer · Jul 2020 – Jul 2022

MANGO CONSULTANT

Built multilingual chatbot, automation, and vision platforms for enterprise marketing, support, and operations teams.

  • Launched CRM chatbot pipelines ingesting Facebook + LINE data with KNN segmentation and Naive Bayes spam filtering.
  • Delivered face-recognition attendance powered by DLIB, CMake, and GPU-accelerated inference.
  • Designed microservices for LINE Developer quotation flows, backed by event-driven messaging and RPA bots.
PythonFastAPITensorFlowDLIBDockerLINE Messaging API

2023 – Present

Principal Builder · 2023 – Present

FREELANCE

Design, build, and launch AI-powered products for hospitality, education, and climate technology founders.

  • Serverless booking platform with payments, automations, and BI.
  • Launched e-learning ecosystem with async workflows and analytics.
  • Consulted on AI platform roadmaps, governance, and talent enablement.
Next.jsSupabaseAWSn8nVercelStripe/OPN
Outcome-Based Skills

Capabilities grouped by the business outcome they support.

No rating bars. These groups describe how the stack turns data, models, cloud services, and automations into production systems.

AI / ML Engineering

Turns raw behavior, spatial, and operational data into model-backed decisions.

Python ML services

Builds reliable Python services and data jobs around model workflows.

Model workflow design

Defines features, labels, inference paths, and review loops around business decisions.

Batch inference

Creates repeatable batch inference flows for campaign, loyalty, and operations cadence.

MLOps / ML Platform

Makes models deployable, traceable, monitored, and easier to improve.

MLflow lifecycle

Uses tracking, registry, and experiment context to keep model work reproducible.

Azure ML / SageMaker

Works with Azure ML and SageMaker patterns for training, jobs, and inference.

Monitoring patterns

Designs drift, quality, and operational monitoring around model outputs.

LLM / Agent Systems

Creates governed language-model systems that can answer, retrieve, and act safely.

RAG systems

Connects approved knowledge sources to LLM interfaces with retrieval workflows.

Agent/tool contracts

Defines tool contracts, prompt behavior, and integration boundaries for agentic systems.

Governed LLM access

Balances usability with identity, policy, evaluation, and audit requirements.

Data Engineering

Shapes messy operational data into features, summaries, and analytics-ready models.

Feature pipelines

Builds feature pipelines for segmentation, churn signals, rewards, and brand health.

Geospatial processing

Processes raster, vector, satellite, drone, and ground-truth data for spatial analysis.

Warehouse-ready models

Designs data models that support dashboards, summaries, and downstream automation.

Cloud Architecture

Creates reliable AWS and Azure foundations for data products and AI systems.

AWS / Azure systems

Builds across AWS and Azure with service choices matched to workload cadence.

Serverless architecture

Uses event-driven and serverless patterns when they reduce operational load.

Infrastructure patterns

Keeps deployment patterns explicit, repeatable, and observable.

Full-stack Product Engineering

Turns AI and automation work into usable products for operators and customers.

Next.js products

Builds Next.js interfaces that make data products fast and understandable.

API and data contracts

Defines API and data contracts that keep UI, backend, and automation aligned.

Business UX

Designs business workflows around the decisions users need to make.

Automation & Observability

Closes the loop from insight to action with workflows, alerts, and monitoring.

n8n workflows

Uses n8n and custom jobs to connect events, notifications, reports, and operations.

Operational alerts

Creates operational alerts that make failures and decision changes visible.

Traceable delivery

Keeps systems traceable from data source to action so teams can debug outcomes.

Certifications

Creds that back the craft.

AWS credentials, beta programs, and continuous education that underpin AI, geospatial, and cloud delivery.

Issued Nov 2024

AWS Certified AI Practitioner

Amazon Web Services

Validates generative AI foundations, responsible AI practices, and solution deployment patterns across AWS and open tooling.

Loading badge…

Issued Nov 2024

AWS Certified AI Practitioner — Early Adopter

Amazon Web Services

Recognises participation in the beta cohort with applied generative AI labs, evaluation frameworks, and governance accelerators.

Loading badge…

Issued Nov 2024

AWS Certified Cloud Practitioner

Amazon Web Services

Demonstrates cloud fluency across architecture, security, billing, and foundational AWS services that underpin delivery.

Loading badge…

Continuing Education & Recognition

  • AWS Security Engineering on AWS

    BSI Training Academy · Facilitated by Udosak Suntithikavong

    Aug 2022 – Nov 2023

  • Developing on AWS

    Amazon Web Services

    Aug 2022 – Nov 2023

  • AWS Technical Essentials

    Amazon Web Services

    Aug 2022 – Nov 2023

  • AWS Cloud Practitioner Essentials (Classroom Day)

    Amazon Web Services

    Aug 2022 – Nov 2023

  • Voxy Proficiency Achievement Certificate – Intermediate

    Voxy

    Nov 2023

  • National Research Office Inventor's Day Presenter

    Thailand National Research Office

    Feb 2020

  • SAU English Test Course

    Southeast Asia University

    Mar 2020

  • Information Technology Contest of Thailand — Great Commission (19th Cohort)

    National Science and Technology Development Agency

    Aug 2019 – Mar 2020

  • NECTEC Innovation Programme Contributor

    National Electronics and Computer Technology Center (NECTEC)

    Mar 2019

FAQ

Direct answers for AI search and human readers.

Short answers about Kane, his current work, and the systems represented in this portfolio.

Who is Watcharapon “Kane” Weeraborirak?

Watcharapon “Kane” Weeraborirak is an AI Engineer based in Bangkok who builds production intelligence systems across ML pipelines, LLM platforms, data products, and automation.

What does Kane specialize in?

Kane specializes in AI/ML engineering, MLOps, LLM and agent systems, data engineering, cloud architecture, full-stack product engineering, and workflow automation.

Where does Kane work now?

Kane currently works at Buzzebees, building AI systems for loyalty and customer engagement platforms.

What AI systems has Kane built?

Kane has built brand health intelligence systems, Goatie for Thai SME accounting workflows, enterprise GenAI platforms, geospatial carbon and forest analytics, booking automations, and learning platform automation.

What is Goatie?

Goatie is an AI accounting assistant for Thai SMEs that helps business owners issue documents, scan receipts, organize expense evidence, and understand income and expenses through a LINE-first workflow.

Does Goatie replace an accountant?

No. Goatie helps prepare and organize documents, but professional tax and accounting review may still be needed.

What is Brand Health Intelligence?

Brand Health Intelligence is a system for turning loyalty, campaign, engagement, merchant, reward, and customer signals into business summaries and marketing actions.

What technologies does Kane use?

Kane uses Python, TypeScript, Next.js, Azure ML, MLflow, Databricks, Azure Blob, AWS, SageMaker, Bedrock, PostGIS, LangChain, n8n, and observability tooling.

Contact

Let’s build the next milestone.

Introduce the challenge, the target metrics, and the timeframe. I’ll respond within one business day.

Sharp ideas deserve precise execution. Share context upfront—business outcomes, technical constraints, non-negotiables—and I’ll propose a roadmap that balances speed with operational calm.

Email KaneSchedule intro