Fachrul Kurniansyah

Data & AI · Bali, Indonesia

Hi, I'm Fachrul

Most analysts read the data. I build the systems that make it exist, flow, and mean something.

Driving analytics, automation, and AI in hospitality.

About Me

Numbers and logic just made sense to me growing up. I studied Statistics at Universitas Brawijaya — the computing-heavy track — so I spent a lot of time with R, Java, and C before I touched real data.

When I started my first job and picked up Python, something shifted. I was solving problems that mattered to real people, not just coursework. That's when I stopped thinking of it as a career and started thinking of it as a craft.

I like building things. A pipeline that runs cleanly every morning, a dashboard someone opens before their coffee, a model that changes what decision gets made — that's what keeps me interested.

Work Experience

Assistant Manager Data & AI @ AYANA Hospitality
Apr 2025 – Present
Data Analyst @ AYANA Hospitality
Nov 2019 – Apr 2025
Part-Time Instructor @ Hacktiv8 Indonesia
Oct 2021 – Jan 2022
Lead Data Scientist @ PT Indonesia Indicator
Oct 2017 – Nov 2019

Skills & Tools

Languages & Libraries

PythonSQLpandasSQLAlchemyPydantic

Databases

PostgreSQLMySQLSQL ServerMongoDBOracleRedis

AI & LLM

OpenAIAnthropicLangChainLangSmithWeaviate

Infrastructure & Tools

Apache AirflowMetabaseDockerGitGitHub ActionsFastAPI

Things I've Shipped

Enterprise BI Platform

Live

Metabase for 300+ people across Sales, Revenue, Finance, F&B, Spa, Banquet, Guest Recognition, and other operational teams. 200+ dashboards, ~100 daily active users. If a dashboard breaks, someone notices within the hour. Currently migrating the whole thing off MSSQL.

MetabasePostgreSQLMSSQLSQL

Data Pipeline Infrastructure

Live

50+ Airflow DAGs pulling from 15+ source systems — Oracle PMS & POS, SAP, spa, loyalty, booking engine, guest surveys, Google Ads, and more. Processing millions of records daily. Includes scheduling, monitoring, and failure recovery — keeping the data flowing across the whole operation.

Apache AirflowPythonOracleMySQLPostgreSQLSQL Server

Guest Chat AI

Live

RAG-based chat assistant deployed across 5 properties — answers guest questions about rooms, amenities, dining, and services in real time. ~1,500 conversations/month with up to 98% accuracy. Backed by a property knowledge base in Vector DB, served via OpenAI.

OpenAIWeaviateLangChainPythonFastAPIPostgreSQL

Website Booking AI Agent

Live

Conversational AI agent on the hotel website — helps visitors explore properties, compare options, understand promos, and navigate to booking. ~1,200 conversations/month with up to 95% accuracy and a 7% conversion rate. Handles pre-booking questions that used to go to human support.

OpenAIWeaviateLangChainPythonFastAPIPostgreSQL

Multilingual Ops Chat Translator

Live

Guests message concierge in their native language — Japanese, Chinese, Korean, or English — through the hotel app. The ops team sees everything in English on their dashboard. ~70K messages/month. No language barrier, no delay.

OpenAIPythonFastAPIPostgreSQL

AI Analytics Enrichment Pipeline

Live

An Airflow-integrated pipeline that runs AI daily over guest feedback — categorizing complaints, summarizing stays, inferring preferences, and surfacing activity recommendations. Processes thousands of records daily. Turns unstructured text into structured insight.

OpenAIApache AirflowPythonSQL Server

Occupancy Forecasting Model

Live

A model to forecast hotel occupancy from historical booking data. Current MAE of 16 room nights at D-14. The output feeds directly into daily worker hiring and revenue pricing decisions — so getting it right actually matters.

PythonpandasOracle

Guest Data ETL & Segmentation

Live

Years of guest profile data, cleaned and deduplicated. Then segmented by behavior to support marketing and the loyalty program. Turned a mess of raw records into something the business could actually act on.

PythonpandasOracleMySQLPostgreSQL

Personal Financial Tracker

Live

My own thing — Google Sheets → PostgreSQL → Metabase, running on this server. Built it because I wanted a clean view of my finances without paying for another SaaS tool.

PythonPostgreSQLMetabaseGoogle SheetsDocker

Let's Connect

If something here is interesting to you — a project, a question, or just a conversation about data and AI — feel free to reach out.