Snowflake After-Hours at ANYbotics: Snowflake Native Engineering in Action

Name: Snowflake After-Hours at ANYbotics: Snowflake Native Engineering in Action
Start: 2026-03-24T18:00:00+01:00
End: 2026-03-24T20:00:00+01:00
Location: ANYbotics AG

Swiss German Region

AI data products, Cortex, Dynamic Tables & containerized dlt — all running fully inside Snowflake.

Mar 24, 5:00 – 7:00 PM (UTC)

About this event

Join our Snowflake User Group meetup for an evening of modern, Snowflake-native engineering — proudly hosted at ANYbotics.
This event features two real-world deep dives that showcase what’s possible when you fully leverage the Snowflake platform:
The Swiss Data Atlas
See how a live, AI-enriched database of 730,000+ Swiss companies was built entirely inside Snowflake. Using Cortex for multilingual translation and industry classification, Dynamic Tables for incremental freshness, and Data Exchange & Marketplace for secure distribution — this session walks through the full architecture from raw ingestion to published data product.
Running dlt Inside Snowflake
Learn how to run containerized dlt pipelines using Snowpark Container Services (SPCS), eliminating external infrastructure and reducing compute costs to a fraction of traditional warehouses. From Docker setup to orchestration patterns, this talk provides a practical blueprint for Snowflake-native ELT.
Expect live demos, architectural insights, and actionable patterns for building scalable, cost-efficient data systems.
Perfect for data engineers, architects, and anyone building serious workloads on Snowflake.

When

Tuesday, March 24, 2026
5:00 PM – 7:00 PM (UTC)

Where

ANYbotics AG
83a Hagenholzstrasse
Zürich 8050

Agenda

Welcome & Opening	Arrival, Welcome and Introduction to ANYbotics.
AI_CLASSIFY, AI_TRANSLATE, Dynamic Tables — One Pipeline, 730K+ Swiss Companies	Switzerland's commercial register is public, but using it is painful. The data is multilingual, unclassified, and scattered — making it nearly useless for analytics out of the box. In this session, we show how we built our Swiss Data Atlas as a fully Snowflake-native data product that does the heavy lifting automatically. Using Snowflake Cortex, we translate company descriptions across all Swiss languages and classify every company into NOGA industry codes. Dynamic Tables keep everything fresh incrementally and Snowflake Data Exchange and Marketplace distribute the enriched dataset securely to consumers without copying a single byte. The result: a live, queryable, AI-enriched database of 730,000+ Swiss companies — industry-classified, language-normalized, and ready to use. We'll walk through the full architecture and show it running live, from raw registry ingestion to published data product. If you've ever wondered what a Snowflake data product on the Marketplace can look like end to end — this is for you.
Running dlt Inside Snowflake: A Journey from UDF to SPCS	This talk explores a practical approach to running dlt (data load tool) pipelines directly inside Snowflake using Snowpark Container Services (SPCS), eliminating the need for external infrastructure while dramatically reducing compute costs. After an initial failed attempt to run dlt in Snowflake UDFs, this session demonstrates how SPCS provides the perfect environment for containerized dlt pipelines. You'll learn how to configure an end-to-end dlt pipeline locally, containerize it with Docker, and deploy it to SPCS - where it runs at just 6% of the cost of the smallest warehouse. Key takeaways: Architecture: How to structure dlt pipelines for containerization, including REST API configuration, incremental loading, and custom field handling Container setup: Creating Docker images from dlt projects, configuring Snowflake image repositories, compute pools, and external access integrations Cost efficiency: Running data pipelines at a fraction of traditional warehouse costs while maintaining the ability to trigger via Snowflake tasks Real-world example: Loading Jira issues and projects data using dlt's REST API framework, with pagination, unnesting, and deduplication Orchestration options: Multiple approaches to triggering downstream tasks—from stream-based detection to direct Python-based task execution from the dlt script This session is ideal for data engineers working with Snowflake who want to modernize their ELT processes, reduce costs, and leverage containerization without managing external infrastructure. The approach demonstrated can be applied to any dlt pipeline, making it a versatile pattern for Snowflake-native data ingestion.
Networking & Apero

Speakers

Martin Seifert

Pro Juventute

CRM & Data Analytics

Benjamin Saladin

bi concepts

CEO

Organizers

Nico Ziegler

Senior Field Marketing Manager

See bio

Yoav Magami

Chapter Co-Lead

See bio

Lukas Leichtle

b.telligent

Chapter Co-Lead

See bio

Benjamin Saladin

bi concepts

Chapter Co-Lead

See bio

Sibel Atasoy Wuersch

ANYbotics

Chapter Co-Lead

See bio

CONTACT US

Snowflake Dev Day (In-Person and Virtual Events)

Snowflake After-Hours at ANYbotics: Snowflake Native Engineering in Action

About this event

When

Where

Agenda

Speakers

Martin Seifert

Benjamin Saladin

Organizers

Nico Ziegler

Yoav Magami

Lukas Leichtle

Benjamin Saladin

Sibel Atasoy Wuersch