Join the MDS Fest slack community to access recordings, engage with speakers, and get updates on MDS Fest ’24
Summary provided by Taylor Brownlow, Head of Product at Count
On Day 2 we had over 20 speakers throughout the day, and topics really spanned the gamut: from hands-on coding sessions to inspiring messages from our panelists on the Women Founders in Data panel.
Here are some of the highlights:
How to win friends and influence people when migrating to dbt ft. Dr. Marielle Dado & Eva Schreyer
Dr. Marielle Dado & Eva Schreyer kicked off day 2 with an honest and thoughtful reflection on the realities of introducing dbt to a less-than-enthusiastic team. Each has gone through their dbt roll-outs, Marielle as an excited data engineer, and Eva as a skeptical analyst. They underscore the need to have not just a technical deployment plan, but also one that is empathetic to the people on your team who need to change in the face of dbt. This is where their most poignant advice comes in: Don’t be a hero, be a steward. The difference, Marielle argues, is that a steward is empathetic toward the concerns of analysts and other members of the team, and is able to find a path that recognizes their strengths and lets them grow into the tool over time. Eva, too, departs with wise words for any analysts who are skeptical of dbt and leaves us feeling like new tool migrations don’t have to be so painful.
Data Dev-neXt: how data as code will change the data developer experience ft. Katie Hindson
Data Dev-X is the next big thing in data. Katie Hindson is Head of Product and Data at Lightdash so she is familiar with the painful workflow behind developing data models in dbt, or what she calls the data developer experience (Data Dev-X). Katie outlines the three biggest problems she sees with Data Dev-X today: we have too few testing environments when developing, we have to switch through too many tools to get tasks done, and it’s too hard to keep things in sync. For each of these problems, Katie expands on what they are doing at Lightdash to address them: an ability to test models in different environments to see downstream impacts more clearly, better context aggregation in automated Slack alerts, and how Lightdash can now support your CI/CD processes. More than anything though, Katie wants us to remember the term Data Dev-X, as it’s a term we’re all going to be increasingly familiar with.
How to build your own Reverse ELT app ft. Hugo Lu
It really is that easy (to build a reverse ETL app). Hugo Lu is an experienced data engineer and is now the founder of Orchestra, a no-code orchestration and observability tool for data teams. His talk is a masterclass in the technical ins and outs of building your own reverse ETL app from scratch. He walks through the process, from high-level design, to live demos using Snowflake and BigQuery. If you find yourself having to (or even having to think about building your own reverse ETL app), Hugo’s talk is a must-see.
Women founders in data panel ft. Lindsay Murphy, Mico Yuk, Leah Weiss, Gabi Steele, Stefania Olafsdottir
I knew this one would be a stunner, and I wasn’t wrong. Lindsay’s much-anticipated panel of Women Founders in Data was engaging, honest, vulnerable, and inspiring. Leah and Gabi, co-founders of Preql, Steph, founder of Avo, and Mico Yuk, founder of BI Brainz and head of community and Data.Hub, talk in-depth about their experiences becoming founders of data companies, the tough realities of people still seeing you as a ‘female’ founder and not just a founder, and what we can do to fix the problem of underrepresentation of women and minorities in data leadership teams. This talk is truly for everyone - whether you’re a woman tired of banging on a glass ceiling, someone wanting to know what you can do to make a difference, or someone looking to learn from wise voices in our industry. Please, give this one a listen.
The power of radical transparency ft. Taylor Brownlow
Ah awkward, this was my talk, so I will breeze through this summary quickly. At Count, we’ve been on a journey from a data notebook to a data canvas, and this talk is a summary of how we made that transition and what we learned along the way. In particular, we noticed that our biggest problems in data right now aren’t technical, but people problems (e.g. how do we translate our powerful data stacks into business value?, or how can we stop being task rabbits for the business?). We also noticed that teams who navigated these problems skillfully employed what we call radical transparency: a practice of being open with the business rather than closing them off behind a wall. It was this recognition that led us to see we needed to build a tool that enabled data teams to be more transparent, and open with their business partners. And thus, the canvas was born.
GAAP for data and the SOMA standard ft. Abhi Sivasailam
Abhi Sivasailam, CEO of Levers Labs, and creator of SOMA (Standard Operating Metrics & Analytics) has perhaps thought more deeply about metrics than anyone else in our industry. His talk walks us through the foundations of SOMA, the ambitious project designed to standardize and streamline how companies measure success - to every data team’s benefit.
Data modeling surround sound: are we hearing the same language? ft. Jerrie Kumalah and Phoenix Millacy Jay
Sometimes it can feel like the only thing Analytics Engineers have in common is that we use dbt. Jerrie Kumalah and Phoenix Millacy Jay set out to create a common framework and way of talking about analytics engineering outside of the tools we use (and how we use them). Jerrie and Phoenix propose a 5-step process for data modeling that brings much-needed attention to the world outside dbt into things like how we get requirements, how we engage multiple stakeholders, how we think about technical debt, and scaling beyond this model at this moment. Rather than sticking just to theory, the two give practical advice on how to start having these conversations with your team.
Never be the Bottleneck: Data Engineering and DevEx ft. Dennis Hume
“To be a successful data engineer you have to allow the other roles on the data team to be as successful as possible.”
Dennis Hume, Data Engineer turned DevEx Engineer challenges the notion that data engineers don’t have users. Dennis reminds us that we have many people who depend on our work, and it's our job to think more deeply about how we enable those around us with what we do. He then breaks this down into advice for teams at various stages (from micro data teams) to large-scale enterprises putting in the best-in-class DevOps tools. His advice is both practical (how to approach documentation), and inspirational (how to prioritize empowerment). If you and your team are looking to make a bigger impact on the teams around you, Dennis’s talk has many practical tips to help get you started.
Accelerating Data Stack Maturity at Orrum by Noel Gomez
When Orrum was looking to update their data management with dbt and Snowflake, they faced a challenge around building a platform that would set them up for long term success and which would not create technical debt from day one.
In this session, Noel explores how Orrum set up Snowflake, dbt and implemented best practices from the start.
User story maps for data teams, projects and people ft. Jeff Sloan
Data Teams can easily get stuck in the "Service Trap" of ad hoc analysis, incoming tickets, and fixing broken pipelines. However, Modern Data Teams know there has to be a better way.
They need to understand their users and develop solutions with real leverage. They need to act a lot more like ... product teams! But in the fog of war -- how can you gain this level of focus?
In this session, Jeff breaks down User Story Mapping -- a flexible product management technique that you can use to:
- Identify where your data team can make the biggest impact for your company
- Slice projects into shippable, valuable increments
- Identify opportunities to boost your career (really!)
Where do product analytics fit in the MDS? ft. Rachel Herrera, Elena Dyachkova, Chetan Sharma, Timo Dechau
Until recently, Product Analytics was largely done on legacy platforms that don’t easily integrate into the modern data stack - but over the past couple of years, we’ve seen the emergence of warehouse-native Product Analytics platforms.
The panel explores:
- What are the particular advantages of warehouse-native architecture for Product Analytics?
- What are the disadvantages?
- How does Product Analytics look different in a warehouse-native world? What are the best practices?
Gentle vs Hardcore Data Engineer ft. Dmitry Anoshin
The data engineer's role is very important and critical. What skills should he have, how well should he know the code, algorithms, and data science? Dmitry was able to identify 2 types of data engineers — Gentle and Hardcore. You'll learn about them during this session and decide which one fits you and how to grow your career.
Access Day 1 recordings here