Astrolabe ML Infrastructure Platform

Raden

Raden Tonev

SUMMARY

During a focused three-month engagement with Astrolabe, I designed their AI infrastructure platform from the ground up. With many ideas initially on the table, I helped the founders shape their product’s strategy and vision to ensure alignment with user needs and business objectives. Alongside crafting the core user experience, I also established a cohesive and compelling brand identity, positioning Astrolabe effectively within its market. This project emphasized both strategic insight and meticulous design execution.

THE CHALLENGE

RESEARCH & ANALYSIS

I began my design process by conducting in-depth interviews with Astrolabe’s founders, helping us understand the company’s vision and core identity. Through careful mapping of existing user journeys, I identified critical process steps and pain points, ensuring the design directly addressed user challenges. Additionally, constructing detailed user personas allowed us to keep user needs at the forefront of every design decision.

User Journeys: Creating & Using Models

After establishing the different personas, I mapped comprehensive user journeys to clearly define the end-to-end experiences for both creating and using machine learning models. This approach allowed us to identify key pain points and opportunities, ensuring the platform design directly aligned with users’ workflows and expectations. Visualizing these journeys helped guide strategic decisions, prioritize features, and establish a shared understanding among stakeholders.
Astrolabe User Journeys
Astrolabe User Journeys

STRATEGY

I collaborated closely with Astrolabe to shape their product strategy, identifying key market opportunities, monetization strategies, and essential functionalities. By deeply understanding user challenges such as the difficulty of finding quality datasets and navigating complex workflows, I developed a strategy focused on streamlining the user experience. This included establishing clear monetization pathways that benefit both model and dataset creators, alongside creating a marketplace and intuitive tools that empower users at every technical level.
📈 OPPORTUNITY
Finding quality datasets is challenging because many are incomplete
Special-purpose models are also hard to come by
The workflows for creating and managing models have steep learning curves
💰 MONETIZATION
Model creators earn money from their published models
Dataset creators earn money from their published datasets when others use them
The Astrolabe platform receives a portion of these revenue sources
💡 FUNCTIONALITY
Guided workflow for training, deploying, and managing models
Ability to “look under the hood” and customize models with code
Marketplace for both datasets and models

DESIGN

After defining clear user journeys and aligning on product strategy, I was able to begin designing. I needed to create all key screens for Astrolabe’s platform. My goal was to produce an intuitive, cohesive interface that simplified complex workflows and made powerful features accessible. Each design decision prioritized clarity, ease-of-use, and flexibility, ensuring both novice and advanced users felt confident and empowered.

MODEL MARKETPLACE

The Marketplace provides a user-friendly hub where creators can monetize their models and datasets, and users can easily discover and purchase specialized resources. The intuitive design highlights valuable content, facilitating seamless transactions.

OUTCOME

I guided the founders through design and product strategy, delivered the platform’s complete end-to-end design, and ultimately helped the company secure its seed funding.
Like this project

Posted Jun 6, 2025

Designed Astrolabe's Machine Learning infrastructure platform. I worked with the founding team to define their product strategy, branding, and design their MVP