Skip to content
01Index

Muhammad
Hafizh Fayiz

Data EngineerProduct Ops AnalystKarawang, ID — Open to remote

I build the pipelines that turn scattered ops data into decisions.

Get in touch→View case studies
PIPELINES SHIPPED12+ETL workflows in production
DEPARTMENTS UNIFIED5Cross-team data consolidated
CURRENT GPA3.54Bachelor of Digital Business
APICSVETLLAKEDBBI
02About

I’m a final-year Digital Business student at Universitas Pakuan, currently working as a Product Operations Analyst at PT Boer Technology. My day-to-day is centralizing data that lives in spreadsheets, third-party APIs, and the heads of project managers into a single queryable platform.

The work is half engineering, half translation. I write Python and Prefect flows that move data from A to B reliably, and I write data dictionaries and SOPs so the next person doesn’t have to guess what status_2 means.

Outside of pipelines, I dig into product analytics — RFM segmentation, retention curves, the kind of analysis that tells a sales lead which 200 customers to call this week.

“Half engineering, half translation.”

03Experience

  1. JUN 2025 — PRESENT

    Product Operations Analyst

    PT Boer Technology — Remote

    Building the data platform that lets every department at Boer see the same numbers.

    • Engineered automated ETL pipelines in Python + Prefect, extracting from heterogeneous sources, transforming with complex business logic, and loading into MinIO (S3-compatible object storage) with high availability and integrity.
    • Consolidated 5 cross-departmental data domains — PMO KPI/SLA tracking, Managed Services alert monitoring, General Affairs inventory, EduTech platform sales, and marketing performance — into a unified Metabase reporting layer.
    • Drove RFM (Recency / Frequency / Monetary) segmentation on EduTech customer data, paired with Amplitude’s Mastering Retention playbook, to surface actionable cohorts for the growth team.
    • Partnered with project managers to translate raw metrics into executive-ready narratives, sharpening KPI clarity in weekly leadership reviews.
    • Authored and maintained the team’s data dictionary and SOP library, reducing onboarding friction for new analysts.
    • Python
    • Pandas
    • Prefect
    • MinIO
    • Metabase
    • DuckDB
    • PostgreSQL
    • Docker
    • Linux

  2. FEB 2025 — JUN 2025

    Intern

    PT Boer Technology — Remote

    Learned the stack by building it from the ground up.

    • Built foundational ETL pipelines pulling from external REST APIs, transforming with Pandas, and orchestrating via Prefect tasks, flows, and deployments.
    • Containerized the analytical environment using Docker (Dockerfile + Compose) for reproducible local-to-server parity.
    • Stood up the local analytical stack: MinIO for object storage, DuckDB as the query engine, Metabase as the visualization layer.
    • Completed a Linux Administration course on the ADINUSA platform.
    • Tracked deliverables and decisions in GitLab issue tickets, keeping the project audit trail tight.
    • Python
    • Pandas
    • Prefect
    • Docker
    • MinIO
    • DuckDB
    • Metabase
    • GitLab
04Selected Work

From Zero to Local Data Stack

Building a reproducible MinIO + DuckDB + Metabase environment that runs identically on a laptop and on the team server.

2025 · Infrastructure · Docker · 2025→
RFM segmentation scatter plotLearners plotted by recency against frequency, point size by monetary value, coloured by segment: Champion, At-Risk, and Hibernating.RECENCY →FREQUENCY →
RFM data points
SegmentRecencyFrequencyMonetary
Champion924138
Champion883647
Champion792922
Champion712431
At-Risk581917
At-Risk492226
At-Risk441412
At-Risk371721
Hibernating2697
Hibernating19611
Hibernating1345
Hibernating879

RFM Segmentation for EduTech Retention

Segmenting learners into Champions, At-Risk, and Hibernating cohorts to drive a targeted re-engagement campaign.

2025 · Product Analytics · Retention · 2025→

Unified Ops Data Platform

How we collapsed five departments' worth of fragmented spreadsheets into one queryable platform.

2025 · Data Platform · ETL · 2025→
05Stack
Data stack flow diagramFour sources — APIs, spreadsheets, PostgreSQL, and webhooks — flow into Prefect for orchestration, then into MinIO object storage, then DuckDB for querying, and finally Metabase for dashboards.APIsSpreadsheetsPostgreSQLWebhooksPrefectMinIODuckDBMetabase

Hover or focus a node to see how I use it.

Data stack pipeline stages
StageToolPurpose
SourceAPIsExternal REST endpoints — partner platforms and SaaS tools.
SourceSpreadsheetsManual ops trackers maintained by project managers.
SourcePostgreSQLTransactional product databases.
SourceWebhooksEvent payloads pushed from third-party services.
PipelinePrefectOrchestrates extract + transform as tasks, flows, and deployments.
PipelineMinIOS3-compatible object storage — the raw + processed data lake.
PipelineDuckDBIn-process analytical query engine over the lake.
PipelineMetabaseBI layer — dashboards every department reads from.

Every pipeline I ship lives somewhere on this diagram. Hover a node to see how I use it.

06Skills
Skills matrix by category and proficiency
CategoryDailyComfortableLearning
LanguagesPython, SQLBashGo
OrchestrationPrefect—Airflow, Dagster
StorageMinIO, PostgreSQL, DuckDB—Iceberg, Delta Lake
Analytics & BIMetabase, PandasAmplitudedbt
InfraDocker, Git, LinuxDocker ComposeKubernetes, Terraform
MethodsETL design, RFM, Data modelingRetention analysisStreaming, CDC
  • Technical Communication
  • Analytical Thinking
  • Problem Solving
  • Cross-functional Collaboration
  • Technical Writing
07Contact

Let’s build something queryable.

Open to data engineering, analytics engineering, and product-ops roles. Remote-first or Indonesia-based.

LinkedInlinkedin.com/in/hafizcareer↗CV (PDF)/cv.pdfdownload

Last updated · 2026-05-20