David Badovinac

A data consultant turns the data you already collect into answers you can act on. In practice that means finding where your data lives, connecting it into one place, defining the metrics everyone agrees on and building the dashboards or tools people use day to day. If your team spends hours stitching spreadsheets together, argues about whose numbers are right, or cannot answer basic questions about revenue, customers or operations without asking an engineer, a data analytics consultant will usually pay for themselves quickly. I start with the decisions you need to make, not the tools: audit where your data lives (app databases, spreadsheets, SaaS tools like Stripe, HubSpot or Shopify), agree the handful of metrics that matter, and build the shortest path to a trusted dashboard.

A modern data stack is a set of cloud tools that work together: connectors that load data from your apps, a cloud data warehouse such as BigQuery or Snowflake to store it, dbt to clean and model it, and a BI tool such as Looker for dashboards. Each piece is managed and pay-as-you-go, so a small team can run analytics that used to need a big IT department. If your reporting lives in spreadsheets or exports from individual tools, a lean modern data stack is usually the step that fixes it. It is also the foundation any serious AI project needs.

A data engineer builds the pipelines that move data into the warehouse. An analytics engineer turns that raw data into clean, tested, documented models (usually with dbt) and defines the metrics the business reports on, sitting between engineering and analysis. I work across both, which is why one person can take you from scattered sources to dashboards and AI tools you can trust.

Yes. As a data engineering consultant I design and build data pipelines that move data from your sources into BigQuery or Snowflake, then transform it with dbt into clean, tested, documented tables. Pipelines are scheduled and monitored, so reports refresh on their own and you hear about broken data before your stakeholders do.

A data warehouse consultant designs the central database that all your reporting and AI runs on: choosing the platform, modelling the data, building the pipelines that load it and keeping costs and performance under control. If you join data from more than one system, or your reports slow down your production database, you need one. BigQuery suits teams already on Google Cloud and spiky workloads; Snowflake suits multi-cloud setups and predictable compute. I help you choose based on your stack, data volume and budget, then set it up.

It depends mostly on how many data sources you have, how messy they are, and how much reporting you need on top. Warehouse running costs on BigQuery or Snowflake are often modest for small and mid-sized companies when tables are partitioned and queries are well modelled; the bigger cost is the build. After a short discovery call I send a fixed-scope proposal with a clear price, so you know the cost before any work starts.

Yes, this is my most common project. As a Looker consultant I review your LookML, explores and warehouse queries, fix the root causes of slow dashboards (fan-out joins, missing aggregates, unpartitioned tables, poor caching), clean up duplicated content and set up a structure your developers can maintain. I have run Looker environments with 1,000+ users.

Yes. I have led Looker migrations for organizations with 1,000+ users. A migration covers inventorying existing reports, rebuilding the logic once in a governed LookML model, recreating the dashboards people really use, training users and running both tools in parallel until everyone has switched.

An AI implementation consultant takes AI from a demo to something your team uses every day: connecting models to your real data and tools, putting guardrails around them and measuring whether they help. AI is only as good as the data behind it, so I connect LLM assistants and conversational analytics to a governed semantic layer (Looker or dbt metrics). Every answer is calculated from an agreed definition rather than guessed, and you can trace it back to the query.

Recurring reports, data-quality checks, alerting when a metric moves, reconciling data between systems, and repetitive admin such as generating LookML through the Looker API. I build AI agents and automated workflows that run on a schedule and hand people only the exceptions.

Many AI automation agencies chain no-code tools together until something breaks. I build AI automation done properly, the way real developers build software: code in version control, tests, error handling, logging and monitoring, running on your own cloud with your data staying where it belongs. As an AI consultant who is also an analytics engineer, I make sure the automation runs on clean, governed data, so it keeps working after the demo.

Yes. I have taught 16,000+ students analytics engineering and Looker on Udemy, and I run hands-on Looker, LookML and dbt training for client teams, from business users building their first dashboards to developers maintaining the model, so the work stays maintainable after the engagement ends.

Book a 30-minute call or send a message. We talk through your data, your goals and your timeline, and I come back with a scoped proposal: a fixed-scope project for things like a migration or a new pipeline, or an ongoing monthly engagement for continuous data and AI work. You work directly with me as an independent consultant, not a large agency, remotely with companies in the US, UK and Latin America.