Designing a no-code, AI-powered analytics tool to help business teams move faster.
MarkovML • 2024 - 2025 • Feature Development
Introduction
MarkovML’s Data Insights is a no-code, AI-powered analytics tool that helps users explore and understand structured data through a simple conversational interface. Users can ask questions in plain English to get instant answers, generate visual reports, and easily build dashboards, all without needing SQL or technical expertise.
My role
I was the sole designer on this project, working directly with product lead & engineers to shape the experience from the ground up. Together, we explored multiple directions through brainstorming, prototyping, and testing to figure out what would best help users explore data, generate insights, and build dashboards, all through a simple, conversational interface.
My contributions included:
Partnering with Product to shape requirements, and with Engineering to align technical possibilities.
Defining the initial UX strategy
Designing interaction flows and visual design
Prototyping and validating concepts built in demo workspaces, using feedback from live product demos with potential customers
Current Challenge
Traditional BI Tools Are Holding Teams Back
Complex UX with steep learning curves for non-technical users
Need to write SQL or know data structures
Leave many users unsure how to frame the right queries
Slow time-to-insight for business teams
As a result, teams across Marketing, RevOps, and GTM often rely on analysts for even basic queries, slowing down decisions and creating bottlenecks.
Bottlenecks in the Go-To-Market Funnel
We asked ourselves: "How might we design an AI-powered experience that makes it easy for any business user, regardless of technical skill to explore data and uncover insights they can trust?"
The goal was to rethink the data exploration experience from the ground up. We wanted users to simply ask questions in plain English and get meaningful, accurate answers, visual reports, and dashboards they could use right away. Behind the scenes, the AI needed to understand natural queries, write context-aware SQL, and maintain accuracy.
Key UX Solutions
Explore data with a natural language chat interface
Ask questions in plain English to explore your data, no SQL required
See the SQL & a simple explanation, so you always know how insights were generated
Build dashboards through conversation
Pin charts directly from chat without switching tools
Connect multiple data once
Explore multiple datasets, whether uploaded manually or integrated from platforms like Salesforce, HubSpot, or Google Ads. This flexibility is crucial during dashboard creation, where users often need to bring together data from different sources to tell a complete story.
Preview tables to ask better questions
Easy access to Data Preview gives users a quick view at what tables and fields are available before they ask a question. This feature improves how users formulate queries.
Impact & Outcomes
Faster time-to-insight
Teams can from question to insight in minutes instead of hours, enabling faster decision-making across functions.
Enabled Self-Serve Insights
Non-technical users were able to explore data and build dashboards without needing analyst support
Strong Early Adoption
Early adoption across Marketing, RevOps, and Sales teams during pilot phase showing positive signs of market fit.
Positive use feedback
Pilot users appreciated the ease and approachability of the natural language interface
Reflections
Designing this module reinforced for me:
The power of conversational UX in simplifying complex workflows
The importance of building trust in AI-driven tools
How reducing friction can drive adoption among non-technical users
Back to top
