Instructor-Led Snowflake & Databricks Team Training for Analytics & Platform Managers

Your team will be able to: Build a governed Bronze-Silver-Gold pipeline in Databricks — ready for BI and reporting the same day.

Build a Bronze, Silver, and Gold Medallion Architecture pipeline in a real Databricks workspace. Covers Delta Lake, Unity Catalog governance, and Lakeflow orchestration.

Your team will be able to: Your analysts stand up a governed Genie agent, fix its wrong answers with Knowledge Store curation, and leave with a benchmark process to keep it trustworthy.

SQL-only, no Python. LLM fundamentals taught with ai_query(), then hands-on Genie: create a space, break it on purpose, fix it with descriptions, synonyms, example SQL, and instructions — and govern it all through Unity Catalog.

Your team will be able to: Equip your finance team to use Claude for real analyst work — models, memos, and reporting — with the guardrails a finance workflow actually needs.

Hands-on workshop for finance professionals: use Claude for financial modeling, variance analysis, and memo drafting, wire it into spreadsheets and internal data via MCP, and apply guardrails for accuracy, auditability, and confidential data handling.

Your team will be able to: Design and operate the full AWS data stack — raw S3 files to a governed query layer in Athena and Redshift.

Ingest raw data into S3, crawl schema with Glue, transform CSV to Parquet with PySpark ETL, and query with Athena. The foundation of every AWS data platform.

Your team will be able to: Build pipelines that react to data as it lands — no polling, no cron jobs.

Wire S3 events to SQS and Lambda to build an event-driven ingestion pipeline. Three patterns: S3 to SQS to Lambda, direct S3 to Lambda, and Lambda-as-ETL.

Your team will be able to: Enforce row, column, and tag-based access on your data lake without writing custom auth code.

Apply row-level, column-level, and tag-based access control to a live data lake using AWS Lake Formation. Real multi-persona access scenarios.

Your team will be able to: Query S3 data lake files directly from Redshift and run federated queries against live databases.

Stand up Redshift Serverless, query S3 directly with Spectrum, and run federated queries against Aurora PostgreSQL. Covers the analytics tier end-to-end.

Your team will be able to: Replicate row-level database changes to the cloud in real time — full load plus ongoing CDC.

Capture row-level changes from a PostgreSQL source using AWS DMS and land them in S3 or a target database. Covers full load plus ongoing replication.

Your team will be able to: Add full-text search and operational dashboards to your data platform without a third-party SaaS.

Ingest processed data from your data lake into OpenSearch, build search queries, and create operational dashboards. Adds a real-time search and observability layer to the data platform.

Your team will be able to: Your team leaves with a one-page decision map — which Databricks or Snowflake AI tool to use, for which user, for which task — plus a working sense of how these tools connect to Git, MCP, and your existing pipelines instead of operating in isolation.

For data analytics managers whose team has been re-titled from analyst to 'engineer' or 'architect' and handed AI coding agents without the operating model to use them safely. Not a feature tour — a decision map. Covers Databricks Genie, Assistant, and the Mosaic AI Agent Framework alongside Snowflake Cortex Analyst, Copilot, Cortex Agents, and Snowflake Intelligence: what each is, who it's for, and what specifically breaks when it's used for the wrong job. Includes a hands-on look at where these tools have to integrate with Git review, MCP connections, and data ingestion/transformation — because in practice nobody uses one tool in isolation. Part 1 of a two-part series; Part 2 (Snowflake & Databricks AI FinOps) goes deep on cost governance and locked-down MCP integration for teams that have already had a cost or stability incident.

Your team will be able to: Your team leaves with a warehouse cost-governance playbook and a locked-down MCP pattern that lets AI agents query Snowflake or Databricks without blowing up compute, storage, or token spend.

For analytics and platform managers upskilling a team that's already had an MCP integration explode their warehouse cost, or whose company blows through its monthly AI token budget. Covers Snowflake/Databricks cost diagnostics (storage vs. compute, query-level cost analytics, backfill guardrails), least-privilege MCP scoping so AI agents can't run unbounded queries, segregating high-volume chat token spend from precious developer/codegen token spend, model-tier restriction policies, skill/prompt registries for consistent team-wide AI use, and a Grafana-based dashboard pattern that surfaces cost anomalies before month-end instead of after.

Your team will be able to: Run LLM inference in-house with full cost controls and observability dashboards — without depending on managed APIs.

Deploy Qwen3 with vLLM on a GPU instance. Wire in LiteLLM as an API gateway with virtual keys, per-team budget caps, and model routing. Instrument with Prometheus and Grafana dashboards for TTFT and throughput.

Your team will be able to: Write least-privilege IAM policies from scratch and audit existing ones without guessing.

Read, predict, and write IAM policies using a real sandbox policy as the textbook. Six exercises from basic allow/deny logic to least-privilege policy design from scratch.

Your team will be able to: Ship a production-ready full-stack app on AWS — backend, frontend, infrastructure, and CI/CD pipeline.

Build and ship a full-stack application on AWS: React frontend, FastAPI backend, Terraform infrastructure, and a GitHub Actions CI/CD pipeline. Seven chapters covering the complete engineering lifecycle.