Activation resources for data engineers

Snowflake introduced AI Gateway for Advertising, connecting Cortex Agents to Meta and TikTok via Model Context Protocol. It bridges warehouse data like transactions and KPIs with live platform signals to automate cross-channel campaign analysis and optimization.

O Analytics Talks debate como os dados deixaram de ser apenas suporte para se tornarem centrais nas decisões. A discussão foca em engenharia de analytics, métodos de mensuração, pipelines de coleta e transformação de informação bruta em ação prática para sustentar o futuro da análise.

This hands-on workshop demonstrates how analytics agents enable natural language data exploration to bypass manual SQL queries and dashboards. Engineers and data teams explore behavioral patterns, customer segmentation, and how AI agents integrate into modern data analytics pipelines.

Data teams should pivot from passive BI dashboards to Intelligent Decision Support Systems. These architectures combine deterministic pipelines with composable LLM agents for tasks like signal extraction and CRM updates, creating high-leverage workflows with measurable ROI.

Snowflake detailed how unifying disparate data signals across sales and marketing into a single source of truth drove AI ROI. Integrating campaign touches, transcripts, and product signals into connected workflows reduced sales cycles by 25-30% and unified go-to-market data operations.

dlt enables custom reverse ETL pipelines using its @dlt.destination decorator. A data pipeline extracts REST API hiring signals into DuckDB, enriches them via People Data Labs, and syncs records directly to the Attio CRM. It gives data engineers code-level control over destination upsert logic.

Netflix solves recommendation cold-start by concatenating 768-dimensional CLIP image embeddings with asset ID embeddings. This multimodal pipeline pools cross-canvas signals into a unified model, using reward-based weighting and inverse propensity scoring on exploration logs for offline evaluation.

QuintoAndar unified its CDP on ClickHouse Cloud to process 900M monthly events and serve 300M API requests. By replacing a Lambda architecture and using ClickHouse Managed Postgres as an interface, they made real-time analytical data accessible to tools requiring Postgres compatibility.

Musinsa scaled its customer data platform by migrating to ClickHouse Cloud. The data engineering team cut storage costs by 86.5 percent while processing audience segmentations across 16.4 million members. Using ClickPipes simplified real-time data ingestion and eliminated extra pipeline tools.

Snowflake integrates first-party data with Meta ads using the Meta Conversions API skill and Meta ads MCP. This setup lets data teams securely hash and egress PII from Snowflake to Meta, while marketers use Snowflake CoWork to analyze campaign performance.

Reverse ETL moves prepared, trusted data from warehouses back into operational tools like CRMs and marketing platforms. It bridges the gap between analysis and action, ensuring teams work with a unified truth.

Retailers must move beyond data centralization to activation. Fivetran Activations helps teams operationalize warehouse data across CRM, marketing, and support systems. This reduces brittle integrations, ensuring portability matters as much as activation to avoid vendor lock-in.

Agentic analytics aims to move from dashboards to action by automating business workflows. Success requires a robust semantic layer and governance rather than just a chatbot. What dragged 21% to 95% was the scaffolding around the model, governed data, a semantic layer it is forced to consult.

Pinterest is leveraging a decade of taste data to automate the retrieval and selection tasks once handled by analysts. Its new Performance+ model and MCP protocol integrate campaign insights directly into external workflows, shifting the focus from manual data pulling to high-level strategy.

Zepto uses a Dual Sequence ReRanker to balance long-term habits with real-time session intent. The architecture employs separate transformer encoders and target-aware pooling to score candidate SKUs within milliseconds.

Bruin streamlines Salesforce to Snowflake pipelines by automating ingestion into bronze, silver, and gold layers. It uses ingestr assets to manage schema evolution and incremental merges. Silver models should answer "what does this mean?" not just "where did this come from?" to ensure data quality.

Netflix uses a hierarchical architecture to decouple notification frequency from content selection. A slow policy sets a weekly pacing plan via a feature store, while a fast policy executes real-time ranking.

Netflix uses a hierarchical architecture to decouple notification frequency from content selection. A slow policy sets weekly pacing plans to manage long-term fatigue, while a fast policy executes real-time ranking. The true power of this hierarchy lies in decoupling.

Predicting customer churn requires centralizing fragmented signals from Gong, Salesforce, and Zendesk into a warehouse like BigQuery. By unifying transcripts and usage data, teams can build AI agents that identify patterns across sources. The full story is only visible when the data is unified.

Snowflake Data Clean Rooms now features the DCR Agent in public preview, a conversational interface for joining collaborations and running audience overlap analyses.

Modern data activation has evolved from simple warehouse-to-CRM field syncing into complex workflows involving AI agents and governed pipelines. The sync itself is not the whole system, as engineers must now manage lineage, quality checks, and diverse outputs like Slack reports or GitHub PRs.

The provided text focuses on leadership within the Morningstar Retirement Data team, emphasizing cross-functional data initiatives. It highlights the importance of enterprise data engineering in fintech. Lead the Morningstar Retirement Data team and cross-functional data initiatives.

Modern data architecture for retail must evolve into a real-time sensory system using identity resolution and vector databases for unstructured signals. Data architecture ought to be designed as a real-time sensory and predictive system, not just for storage and analysis of past transaction data.

Snowflake Data Clean Rooms version 15.1 introduces enhanced auditing and data management. Users can now use VIEWACTIVITYHISTORY to audit analysis activity in a collaboration. Additionally, the CLEANROOMPUBLICROLE can now delete records from the activation segment table to clean up data.

Fivetran uses its warehouse-native activation tool to automate operational loops. By syncing modeled data from BigQuery into Salesforce and Postgres, they power billing systems and internal apps. The heavy lifting happens in the warehouse, enabling teams to ingest broadly and analyze centrally.

Embedded analytics integrates dashboards and KPIs directly into SaaS products to drive growth. It requires strict per-tenant data isolation and native integration to reduce churn and increase expansion revenue. Embedded analytics in SaaS is what you deliver to your customers as part of your product.

Snowflake Intelligence enables manufacturers to unify fragmented data from connected products into a single source of truth. By blending structured and unstructured data with real-time context, it uses intelligent agents to replace complex dashboards. The team needed answers in seconds.

🎙️ It's About Data — Explores Systems of Action in data engineering, shifting from record and intelligence systems to action-oriented architectures, leveraging AI and data analytics tools.

Pleo leverages Fivetran Activations and HubSpot Association Labels to operationalize Customer 360 data, ensuring governance and accuracy through custom role associations between contacts and companies.

Fivetran introduces Audience Hub, a no-code, warehouse-native segmentation suite enabling marketers to build and activate audiences directly from the data warehouse with advanced features like writeback and data governance.