From our Nimmanhaemin base in Chiang Mai, we take on paid programming engagements that untangle batch ETL, orchestration, and the glue code between your sources and warehouses.
Python and SQL transforms, Airflow-style DAGs, dbt models, Kafka consumers, and the retry/idempotency layers that keep nightly loads from waking your on-call.
Every project is scoped after an inquiry: we read your schemas, SLAs, and existing stack, then propose deliverables and timing — no catalog pricing on this site.
We work in ICT with teams across ASEAN time zones; documentation and handoffs are written for engineers who will operate the pipeline after we ship.
Many of our clients run hybrid stacks — on-prem extracts in Bangkok, cloud warehouses in Singapore — while product teams sit in Chiang Mai. We have walked the Nimmanhaemin corridor between stand-ups and late-night batch windows.
That locality matters when you need someone who understands Thai business hours, holiday load freezes, and the practical latency of cross-border replication.
Upstream vendors change CSV columns without notice. We program defensive parsers, quarantine tables, and alerting that names the offending file — not a generic “job failed”.
Backfills should be boring. We wire partition-aware reruns, explicit dependencies, and runbooks that fit how your data team actually operates.
Like recycled textile board, good ETL is layered: sourced materials, honest joins, and a finish that survives humidity — the operational kind, when disks fill and APIs throttle.
Stabilize failing nightly loads, add tests around critical transforms, and document lineage enough for the next engineer.
Greenfield ingestion from APIs, SFTP drops, or replica DBs into a warehouse you already pay for.
Architecture and code review blocks for teams that want a second set of eyes on DAG design or cost-aware partitioning.
No. We quote after understanding sources, volumes, and operational constraints. The inquiry form is the starting point.
Usually yes — Airflow, Dagster, Prefect, cron-plus-bash, or managed services. We adapt to what your team already runs.
Most delivery is remote pair-programming with written handoffs. On-site working sessions in Chiang Mai can be arranged when useful.