MARA Grid Intelligence

Industry

Energy / Digital Assets

Design Category

Internship / Product Design

Timeline

Jun 2025 - Aug 2025

Team

Ayesha Khan, Elaine Zhang

Impact
40% Faster Site Review
Consolidated power, land, and infrastructure data into one dashboard — cutting average site evaluation cycles from 3 weeks to under 5 days.
12 Sites Modeled
Design sprint covered 12 candidate data center sites across 4 states, producing a repeatable scoring framework for MARA's expansion decisions.

Background

MARA Holdings (previously Marathon Digital Holdings) is one of the largest publicly traded bitcoin mining companies in North America. The company focuses on building and operating high-performance data centers powered by renewable and cost-efficient energy sources. Its mission is to scale bitcoin mining and digital asset compute in a way that is profitable, sustainable, and strategically aligned with the evolving energy landscape. At the core of its business, MARA invests heavily in infrastructure, technology, and partnerships that allow it to operate at global scale while reducing environmental impact. The MARA Grid Intelligence Platform was developed as an internal design sprint concept to explore how teams could improve the way they discover, evaluate, and secure new sites for data centers.

MARA Holdings (previously Marathon Digital Holdings) is one of the largest publicly traded bitcoin mining companies in North America. The company focuses on building and operating high-performance data centers powered by renewable and cost-efficient energy sources. Its mission is to scale bitcoin mining and digital asset compute in a way that is profitable, sustainable, and strategically aligned with the evolving energy landscape. At the core of its business, MARA invests heavily in infrastructure, technology, and partnerships that allow it to operate at global scale while reducing environmental impact. The MARA Grid Intelligence Platform was developed as an internal design sprint concept to explore how teams could improve the way they discover, evaluate, and secure new sites for data centers.

MARA's operating digital asset data center in Granbury, TX, purchased from Generate Capital for $178.6 million in December, 2023.

Problem

How might we give MARA’s teams a smarter, faster way to evaluate and secure sites for digital asset compute?

MARA’s site selection process is critical to its growth, but the way decisions are currently made creates systemic obstacles. Data lives in silos across spreadsheets, reports, and verbal updates, which makes it difficult to align across teams. Scoring methods varied, so a site considered “high potential” by one group could be rejected by another. Executives lack clarity, Finance lacks standardization, and Analysts are buried in manual work. The bigger problem is not just inefficiency, it's that MARA risked missing the right sites altogether because the process was not built for speed, consistency, or trust.

Delays = missed opportunities. Slow and inconsistent site/partner evaluation wastes resources, delays operations, and allows competitors to move faster.

Research

OBJECTIVE

To uncover the pain points that we need to address in our solution, in depth user research is required. The purpose of this research is to understand how MARA’s internal teams (Corporate Development, Operations, Executives, Finance) currently identify, evaluate, and prioritize potential partner sites, and to uncover the bottlenecks, gaps, and opportunities for improvement.

Primary Goals

Document current workflows

Document current workflows

Identify role-specific pain points

Identify role-specific pain points

Determine critical data needs for faster site selection

Determine critical data needs for faster site selection

Validate the feasibility of a centralized intelligence dashboard

Validate the feasibility of a centralized intelligence dashboard

METHODS

8 Stakeholder Interviews

Conducted 8 semi-structured interviews (2 per role: CD, Operations, Executive, Finance) to understand tools, workflows, and pain points.

2 Process Observations

Observed 2 live site evaluation sessions with the CD and Ops teams to see firsthand how data was gathered and decisions were made.

Journey Mapping Workshop

Held a collaborative workshop with 6 team members to map the “current state” journey for evaluating a new lead from discovery → final decision.

Pain Point Clustering

Synthesized notes into an affinity diagram to identify recurring problems and categorize them by workflow stage.

PERSONAS

We documented blockers that happen at various stages of scouting and evaluating a physical site for acquisition or partnership and identified key characteristics of the team members that would interface with our solution.

  • Prioritize high-value leads and filter out low-feasibility opportunities early.

  • Present compelling, data-backed pitches to prospective partners.

  • Build and maintain a steady pipeline of viable projects.

Key Goals

  • Leads tracked in scattered spreadsheets; no unified system for ranking.

  • Time wasted chasing prospects with poor technical or financial fit.

  • Lacks quick access to climate and energy cost data during early conversations.

Pain Points

  • Fast-moving, deadline-driven.

  • Comfortable with digital tools but prefers simple, actionable dashboards.

Behavioral Traits

  • AI-driven lead scoring to help identify top potential sites

  • Integrated location search with climate and grid data overlays.

  • A centralized platform to quickly access climate, energy, infrastructural data.

Opportunities

Corporate Development Manager

Primary User

Workflow Map

Relies on networking, industry events, cold outreach, and scattered public datasets.

Uses multiple tools (Google, LinkedIn, government reports, Excel) to assess viability.

Partner potential judged mostly on anecdotal knowledge, slow manual verification.

Sends raw or incomplete site info to Ops/Finance for further due diligence.

Manual Lead Sourcing

Time intensive

Fragmented Research

Inconsistent data quality

Ad-hoc Evaluations

Difficulty prioritizing best-fit opportunities quickly

Internal Handoff

  • Prioritize high-value leads and filter out low-feasibility opportunities early.

  • Present compelling, data-backed pitches to prospective partners.

  • Build and maintain a steady pipeline of viable projects.

Key Goals

  • Leads tracked in scattered spreadsheets; no unified system for ranking.

  • Time wasted chasing prospects with poor technical or financial fit.

  • Lacks quick access to climate and energy cost data during early conversations.

Pain Points

  • Fast-moving, deadline-driven.

  • Comfortable with digital tools but prefers simple, actionable dashboards.

Behavioral Traits

  • AI-driven lead scoring to help identify top potential sites

  • Integrated location search with climate and grid data overlays.

  • A centralized platform to quickly access climate, energy, infrastructural data.

Opportunities

Corporate Development Manager

Primary User

Workflow Map

Manual Lead Sourcing

Time intensive

Relies on networking, industry events, cold outreach, and scattered public datasets.

Fragmented Research

Inconsistent data quality

Uses multiple tools (Google, LinkedIn, government reports, Excel) to assess viability.

Partner potential judged mostly on anecdotal knowledge, slow manual verification.

Ad-hoc Evaluations

Difficulty prioritizing best-fit opportunities quickly

Sends raw or incomplete site info to Ops/Finance for further due diligence.

Internal Handoff

  • Access accurate, complete datasets without hunting across multiple tools.

  • Apply standardized feasibility scoring to improve decision consistency.

  • Reduce manual work in modeling heating demand, ROI, and CO₂ impact.

Key Goals

  • Pulls data from 5–6 unconnected systems, leading to delays and errors.

  • Lack of integrated GIS tools for visualizing energy needs and infrastructure.

  • Scoring methods vary across analysts, creating inconsistency in evaluations.

Pain Points

  • Highly detail-oriented, data-driven.

  • Comfortable with technical tools but frustrated by repetitive manual tasks.

Behavioral Traits

  • Centralized dashboard with API connections to climate, energy, and cost databases.

  • Customizable scoring models with role-based access.

  • Automated report generation for internal review.

Opportunities

Technical Operations Analyst

Primary User

Workflow Map

Often missing technical or energy data.

Contacts local municipalities, utilities, or partners for specs like infrastructure readiness, heating demand, and regulations.

Expensive and time-consuming trips for technical validation.

Sends findings back to Corp Dev/Finance for decision-making.

Receives Basic Site Info from Corp Dev

s

Manual Data Collection

Lack of centralized, reliable data sources

Ad-hoc Evaluations

High costs for site validation

On-site Visits

Long delays between opportunity identification and feasibility confirmation.

  • Access accurate, complete datasets without hunting across multiple tools.

  • Apply standardized feasibility scoring to improve decision consistency.

  • Reduce manual work in modeling heating demand, ROI, and CO₂ impact.

Key Goals

  • Pulls data from 5–6 unconnected systems, leading to delays and errors.

  • Lack of integrated GIS tools for visualizing energy needs and infrastructure.

  • Scoring methods vary across analysts, creating inconsistency in evaluations.

Pain Points

  • Highly detail-oriented, data-driven.

  • Comfortable with technical tools but frustrated by repetitive manual tasks.

Behavioral Traits

  • Centralized dashboard with API connections to climate, energy, and cost databases.

  • Customizable scoring models with role-based access.

  • Automated report generation for internal review.

Opportunities

Technical Operations Analyst

Primary User

Workflow Map

Receives Basic Site Info from Corp Dev

s

Often missing technical or energy data.

Manual Data Collection

Lack of centralized, reliable data sources

Contacts local municipalities, utilities, or partners for specs like infrastructure readiness, heating demand, and regulations.

Expensive and time-consuming trips for technical validation.

Ad-hoc Evaluations

High costs for site validation

Sends findings back to Corp Dev/Finance for decision-making.

On-site Visits

Long delays between opportunity identification and feasibility confirmation.

  • Receive complete, standardized financial data for all sites under review.

  • Accurately calculate payback periods and cost savings for potential partners.

  • Deliver consistent investment cases to funding committees.

Key Goals

  • Site evaluation data often arrives incomplete or in inconsistent formats.

  • Manual re-formatting for finance reviews wastes hours per project.

  • Lack of integrated tool for calculating and visualizing payback and ROI.

Pain Points

  • Risk-conscious, detail-focused.

  • Relies heavily on consistent metrics and structured reporting.

Behavioral Traits

  • Auto-generated financial summaries based on standardized formulas.

  • Pre-formatted report templates for investment committees.

  • Integration with MARA’s existing finance software.

Opportunities

Financial Analyst

Secondary User

Workflow Map

Often without early access to technical/market data.

Builds ROI models using static spreadsheets and outdated benchmarks.

Hard to model multiple financial outcomes quickly.

Slows down the deal-closing process.

Receives late-stage proposals

Reactive rather than proactive in evaluations.

Manual cost modeling

Limited integration between market, technical, and financial data.

Limited Scenario Testing

Delayed Budget Approval

Missed opportunities due to slow budget sign-off.

  • Receive complete, standardized financial data for all sites under review.

  • Accurately calculate payback periods and cost savings for potential partners.

  • Deliver consistent investment cases to funding committees.

Key Goals

  • Site evaluation data often arrives incomplete or in inconsistent formats.

  • Manual re-formatting for finance reviews wastes hours per project.

  • Lack of integrated tool for calculating and visualizing payback and ROI.

Pain Points

  • Risk-conscious, detail-focused.

  • Relies heavily on consistent metrics and structured reporting.

Behavioral Traits

  • Auto-generated financial summaries based on standardized formulas.

  • Pre-formatted report templates for investment committees.

  • Integration with MARA’s existing finance software.

Opportunities

Financial Analyst

Secondary User

Workflow Map

Receives late-stage proposals

Reactive rather than proactive in evaluations.

Often without early access to technical/market data.

Manual cost modeling

Limited integration between market, technical, and financial data.

Builds ROI models using static spreadsheets and outdated benchmarks.

Hard to model multiple financial outcomes quickly.

Limited Scenario Testing

Slows down the deal-closing process.

Delayed Budget Approval

Missed opportunities due to slow budget sign-off.

  • Quickly understand the viability of opportunities without wading through technical details.

  • Compare ROI, strategic alignment, and impact potential across multiple sites.

  • Make confident, timely decisions to maintain momentum in the pipeline.

Key Goals

  • Overwhelmed by detailed technical reports with inconsistent summaries.

  • Relies on verbal or ad hoc updates from teams; no real-time dashboard access.

  • Difficulty aligning cross-functional teams on priority opportunities.

Pain Points

  • Vision-oriented, time-constrained.

  • Prefers visual summaries over spreadsheets.

Behavioral Traits

  • Executive dashboard view with “Top 5 Opportunities” snapshot.

  • Data visualizations for ROI, cost-benefit, and environmental impact.

  • Exportable summaries for board presentations.

Opportunities

SVP of Corporate Development

Secondary User

Workflow Map

Data comes from multiple departments, often in different formats.

Compares incomplete data to business goals and resource constraints.

Limited visibility into real-time market trends or competitive landscape.

Weeks/months lost before greenlighting projects

Receives Fragmented Reports

No single source of truth for decision-making.

Manual Synthesis

Reliance on partial or outdated data.

Gut-based Decisions

Delayed Execution

Slow approval process reduces competitive advantage.

  • Quickly understand the viability of opportunities without wading through technical details.

  • Compare ROI, strategic alignment, and impact potential across multiple sites.

  • Make confident, timely decisions to maintain momentum in the pipeline.

Key Goals

  • Overwhelmed by detailed technical reports with inconsistent summaries.

  • Relies on verbal or ad hoc updates from teams; no real-time dashboard access.

  • Difficulty aligning cross-functional teams on priority opportunities.

Pain Points

  • Vision-oriented, time-constrained.

  • Prefers visual summaries over spreadsheets.

Behavioral Traits

  • Executive dashboard view with “Top 5 Opportunities” snapshot.

  • Data visualizations for ROI, cost-benefit, and environmental impact.

  • Exportable summaries for board presentations.

Opportunities

SVP of Corporate Development

Secondary User

Workflow Map

Receives Fragmented Reports

No single source of truth for decision-making.

Data comes from multiple departments, often in different formats.

Manual Synthesis

Reliance on partial or outdated data.

Compares incomplete data to business goals and resource constraints.

Limited visibility into real-time market trends or competitive landscape.

Gut-based Decisions

Weeks/months lost before greenlighting projects

Delayed Execution

Slow approval process reduces competitive advantage.

KEY INSIGHTS

Below is a side-by-side visualization showing how the user pain points we identified in research directly translate into functional opportunities for development.

User Pain Points

Data Fragmentation

Site information lives in scattered spreadsheets, reports, and tools, forcing teams to manually piece together data and slowing down analysis.

Data Fragmentation

Site information lives in scattered spreadsheets, reports, and tools, forcing teams to manually piece together data and slowing down analysis.

Inconsistency

Different teams apply varying scoring methods and ROI models, leading to mismatched evaluations and eroding confidence in site decisions.

Inconsistency

Different teams apply varying scoring methods and ROI models, leading to mismatched evaluations and eroding confidence in site decisions.

Delays

Manual validation and rework extend decision cycles, slowing down approvals and reducing MARA’s ability to act quickly on competitive opportunities.

Delays

Manual validation and rework extend decision cycles, slowing down approvals and reducing MARA’s ability to act quickly on competitive opportunities.

Lack of Transparency

Executives and Finance see incomplete or outdated data, while Corp Dev and Ops drown in raw details without a clear, shared view of site potential.

Lack of Transparency

Executives and Finance see incomplete or outdated data, while Corp Dev and Ops drown in raw details without a clear, shared view of site potential.

Feature Opportunitites

Visual Site Comparisons

Data streamlined into charts, maps, and rankings make site comparisons easier, helping executives quickly identify top opportunities and align decisions.

Visual Site Comparisons

Data streamlined into charts, maps, and rankings make site comparisons easier, helping executives quickly identify top opportunities and align decisions.

Standardized Scoring

A unified scoring model ensures consistent evaluations, builds trust across teams, and makes ROI and feasibility assessments clear and comparable.

Standardized Scoring

A unified scoring model ensures consistent evaluations, builds trust across teams, and makes ROI and feasibility assessments clear and comparable.

Faster Feedback Loops

Consolidating data into one platform enables real-time insights, shortens feedback loops, reduces rework, and helps MARA accelerate approvals and act quickly on opportunities.

Faster Feedback Loops

Consolidating data into one platform enables real-time insights, shortens feedback loops, reduces rework, and helps MARA accelerate approvals and act quickly on opportunities.

Data-backed Evaluations

Automated data collection reduces manual work, delivers reliable insights, and ensures Executives, Finance, and Analysts have transparent, up-to-date information for confident decisions.

Data-backed Evaluations

Automated data collection reduces manual work, delivers reliable insights, and ensures Executives, Finance, and Analysts have transparent, up-to-date information for confident decisions.

Ideation

DATA & TECHNICAL FEASIBILITY

How might we explore features like centralized data, automated scoring, GIS integration, and AI insights to solve these challenges?

Before defining features, we evaluated the data landscape and technical feasibility to ensure automated processes and AI integration could be applied responsibly and deliver real, actionable value. Click on each framework below to view in detail.

Site Scoring Framework

Site Scoring Framework

Data Ecosystem

Data Ecosystem

AI Powered

AI Powered

Data Pipeline

Data Pipeline

MOOD BOARD

Having established our data landscape, we then researched how to establish a unique UI direction that balances MARA’s existing brand guidelines with our vision for the platform’s look and feel.

Keywords: Data-centric, clean, tech-forward (AI-driven feel), high-contrast readability.

Geospatial site presentation

Buttons, Cards, Lists, Charts, means of Data Visualization

FEATURE DEVELOPMENT

Using the personas, we mapped out the platform’s core functions through a site map and low-fidelity sketches as a way to structure the user flow, aligning the data-driven requirements with an intuitive UI structure. This step helped us translate complex data inputs into clear workflows.


Site map generated via Relume

Sketches figuring out content placement on the UI

DESIGN TENANTS

Transparency

Every score and AI output must be traceable to trusted, verifiable data sources.

Clarity

Tailor insights to each role, balancing executive snapshots with analyst-level detail.

Speed

Streamline workflows to reduce manual effort and accelerate site evaluation and decision-making.

Trust

Design for confidence in AI-driven recommendations through explainability, consistency, and accountability.

Prototyping

WIREFRAMING

Key Points

  • Mapped core 3 main panels (Home, Geospatial Presentation, Data Transparency) and main user flows (login, to dashboard overview, to detailed site and region-level views).

  • Prioritized clarity over polish to focus on information hierarchy and layout decisions.

Home Page

Home Page

Home Page

Data Transparency Panel

Data Transparency Panel

Data Transparency Panel

Centralized view of external and internal data sources with confidence ratings.

Data Transparency Panel

Home Page

Dashboard (Home)

High-priority sites, KPIs, and candidate counts give teams an instant performance snapshot.

Login

Simple login screen providing secure access.

Data Transparency Panel

Centralized view of external and internal data sources with confidence ratings.

Login

Simple login screen providing secure access.

Dashboard (Home)

High-priority sites, KPIs, and candidate counts give teams an instant performance snapshot.

Data Transparency Panel

Centralized view of external and internal data sources with confidence ratings.

Geospatial Site Presentation

Geospatial Site Presentation

Geospatial Site Presentation

Geospatial Site Presentation

Geospatial Site Presentation

Regional View

Zoomed-in regional view showing latest regional news and shortlisted sites with opportunity and risk summaries.

Global View

Interactive world map with filters for top regions, risks, and incentive zones.


Regional AI Report

AI-generated summary highlighting regional suitability, opportunities, risks, and environmental impact.

Global View

Interactive world map with filters for top regions, risks, and incentive zones.


Regional View

Zoomed-in regional view showing latest regional news and shortlisted sites with opportunity and risk summaries.

Regional AI Report

AI-generated summary highlighting regional suitability, opportunities, risks, and environmental impact.

Site View

Drill-down site page with AI report, ROI scatter plot, and comparison option.

Site-by-Site Comparison

Side-by-side evaluation of two sites with weighted criteria and feasibility breakdowns.

Site View

Drill-down site page with AI report, ROI scatter plot, and comparison option.

Site-by-Site Comparison

Side-by-side evaluation of two sites with weighted criteria and feasibility breakdowns.

Site-by-Site Comparison

Side-by-side evaluation of two sites with weighted criteria and feasibility breakdowns.

DESIGN SYSTEM

Typography

Fonts

MARA’s official typefaces (JetBrains Mono and TTHoves Pro) are used for different instances on the platform.

Scale

The base font scale is 12px for body text, with a 4px increment.

Color

Interface

These are the basic building colors for the background of most interfaces.

States

These are the main colors that hint at states.

Icons

Components

Site Cards

Cards that contain site information at a glance.

Interactive Cards

Cards that users click on, manage, and manipulate.

Charts & Data

Brand aligned, exec-ready data visualization that can be downloaded.

Button States

Primary, Secondary, and Tertiary buttons on the platform and their usage states.

HIGH FIDELITY WALKTHROUGH

ANNOTATED HIGH FIDELITY

Dashboard (Home)

Global Map

Data Transparency

Regional View

Regional View

AI Region Summary Details

Site-by-Site Comparison

Site View

HIGH-FIDELITY HIGHLIGHTS

HIGH-FIDELITY WALKTHROUGH

Dashboard

Home Page

After logging in, users land on the Home Page dashboard, which serves as the central hub for site discovery and portfolio of sites evaluation research. At the top, teams see recommended high priority sites that are high in acquisition value, along with key portfolio metrics underneath that informs them of the overall performance of the teams' portfolio health of site acquisition pipeline.

After logging in, users land on the Home Page dashboard, which serves as the central hub for site discovery and portfolio of sites evaluation research. At the top, teams see recommended high priority sites that are high in acquisition value, along with key portfolio metrics underneath that informs them of the overall performance of the teams' portfolio health of site acquisition pipeline.

Global Map

Core Page

The Global Geospatial View provides a map-based overview of all potential sites worldwide, color-coded by feasibility. It surfaces top opportunity regions, incentive hotspots, and critical risk alerts, allowing teams to zoom in, filter by criteria like energy cost or grid stability, and instantly preview KPIs on hover. This view acts as the entry point for discovery, giving Business Development a way to spot high-potential leads, Analysts quick access to infrastructure context, and Executives a visual sense of where MARA’s strongest opportunities lie.

The Global Geospatial View provides a map-based overview of all potential sites worldwide, color-coded by feasibility. It surfaces top opportunity regions, incentive hotspots, and critical risk alerts, allowing teams to zoom in, filter by criteria like energy cost or grid stability, and instantly preview KPIs on hover. This view acts as the entry point for discovery, giving Business Development a way to spot high-potential leads, Analysts quick access to infrastructure context, and Executives a visual sense of where MARA’s strongest opportunities lie.

Site View

AI Site Report Details

The site view, when expanded, provides and AI-powered site report with feasibility, risks, and key metrics, supported by auto-generated visualizations like a radar chart, financial breakdown, and ROI projection for quick executive reporting. It also includes infrastructure details, ESG impact, and scenario analysis tailored to Ops and Finance. To ensure transparency, a Data Provenance appendix ties every number back to a trusted source and timestamp. Finally, AI-recommended actions guide each team’s next steps, turning insights into clear, actionable decisions.

The site view, when expanded, provides and AI-powered site report with feasibility, risks, and key metrics, supported by auto-generated visualizations like a radar chart, financial breakdown, and ROI projection for quick executive reporting. It also includes infrastructure details, ESG impact, and scenario analysis tailored to Ops and Finance. To ensure transparency, a Data Provenance appendix ties every number back to a trusted source and timestamp. Finally, AI-recommended actions guide each team’s next steps, turning insights into clear, actionable decisions.

IMPACT

Design’s Value

Provided grounds for MARA to explore and implement future-facing tools, helping the organization magine what an AI-enabled decision platform could look like.

Provided a scalable concept

For how MARA could approach site evaluation through a unified, AI-powered platform.

Framework

Tailor insights to each role, balancing executive snapshots with analyst-level detail.

Provided a scalable concept

For how MARA could approach site evaluation through a unified, AI-powered platform.

RETROSPECT

My Key Learnings

Reflection 1

Designing for multiple roles

Aggregating features and solutions that each team needs into a single source for streamlining efficiency.

Reflection 1

Designing for multiple roles

Aggregating features and solutions that each team needs into a single source for streamlining efficiency.

Reflection 2

AI in product design must be explainable

Transparency and provenance are critical to build trust, not just impressive features.

Reflection 2

AI in product design must be explainable

Transparency and provenance are critical to build trust, not just impressive features.

Reflection 3

Grounding design in data feasibility

Avoids concepts that look good in Figma but fail in reality.