Network Analysis Platform Overview

Network Analysis gives you the ability to see, interrogate, and model global supply chains at scale — from the macro picture down to individual entities and back — so you can make informed decisions about risks, dependencies, and opportunities. This guide walks through the core concepts, primary views, and analytical workflows.

What Is Network Analysis?

Network Analysis is built around a macro-to-micro-to-macro analytical model. Start with the big picture: map entire industries and product categories to identify concentration risk, geographic dependencies, chokepoints, and adversary exposure across global supply chains. Drill into progressively finer levels of detail — specific production stages, individual countries and their roles, entity-level ownership structures, and the underlying shipment and transaction data that shape each link in the chain. Then zoom back out to translate entity-level findings into sector-wide assessments, policy-ready recommendations, and strategic responses.

You can also simulate the impacts of trade policies, tariffs, sanctions, and supply disruptions at any level of the analysis — evaluating how a macro-level shock ripples through specific production phases and the entities operating within them. Every metric and simulation result traces back to the underlying data that generated it, ensuring full analytical transparency from the top-level view all the way to individual shipment records.

Who Is This For?

Network Analysis serves three core users in the national security and economic security community:

• Defense Resilience Officers — Map critical defense supply chain networks to surface vulnerabilities (bottlenecks, FOCI risks, adversarial dependencies) and guide procurement decisions in collaboration with the DIB.
• Economic Security Analysts — Surface concentrations, bottlenecks, and adversarial dependencies to shape policy and identify opportunities for strategic investment, new markets, and competitive advantage.
• Trade Enforcement Investigators — Trace goods flows and identify compliance risks across complex, multi-hop supply chains, then pivot to transaction-level evidence.

How Networks Are Created

Networks are generated through a collaborative process:

1. You define the scope: companies of interest, industries (NAICS/NACE codes), or goods (HS codes) relevant to your mission.
2. The Altana team generates multi-tier networks by applying graph exploration tools and LLM-assisted processing to Altana’s global trade intelligence graph.
3. The Altana team performs quality assurance on the network, including automated validation against UN Comtrade data and subject matter expert review in collaboration with you.
4. The Altana team deploys the completed network to your environment, typically within 1–2 business days.

Contact your Altana account team to request a new network. Be as specific as possible about target companies, HS codes, or NAICS codes.

Key Concepts & Glossary

Core Concepts

• Network: A defined set of companies, facilities, and transactions representing a supply chain ecosystem, generated around specific industries, companies, or goods.
• Production Phase: A step in the industrial process derived by LLM analysis of HS codes (e.g., raw material extraction, component manufacturing, final assembly).
• Adjacent Industry: Entities supplying goods necessary for production but not direct components of the primary product.
• Treemap: Visual representation within each production phase.
• Leaf (part of the treemap): a company-owned facility, sized by a selected metric.
• HS Code: Harmonized System code classifying traded goods at 2-digit (broad), 4-digit (heading), and 6-digit (subheading) levels.

Metrics

• Connectivity Score: Total distinct companies an entity interacts with (inbound + outbound).
• Node Degree: Raw count of incoming and outgoing edges on a node.
• Neighborhood Overlap: How much two connected nodes’ neighborhoods overlap (0 = bridge, 1 = tightly interconnected).
• Clustering Coefficient: How tightly clustered a node’s neighborhood is (0–1 scale).

Risk & Ownership

• GUO (Global Ultimate Owner): The ultimate parent company in an ownership structure. GUO country indicates where corporate control resides.
• Exposure: Risk attributes associated with an entity (sanctions, trade restrictions, enforcement actions).
• Country Risk: Derived from World Bank data and organizational membership (OECD, G7). Customizable to specific risk taxonomies.
• Foreign Control: Entities where the GUO country differs from the registration country.

Resilience Terms

• Single Point of Failure (SPOF): A node whose failure disrupts the supply chain with no existing alternative for the same goods.
• Edge-Switch Probability: Scenario parameter controlling recovery attempt speed.
• Robustness: Network’s ability to withstand initial disruption impact.
• Resilience: Network’s ability to recover over time through adaptation and alternative sourcing.

Navigation Overview

The application follows a progressive drill-down structure:

• Landing Page: Select a network to analyze.
• Network Overview: Characterize the network’s composition, concentrations, and risks.
• Flow View: Visualize the industrial production process and explore entities within each phase.
• Map View: See the geographic distribution of entities.
• Global Trade View: Understand macro trade flows using UN Comtrade data.
• Resilience Analysis: Simulate disruptions and assess supply chain robustness.
• Chat: Query the network using conversational AI.
• Pivot to Knowledge Graph: Transfer entities for deep-dive, transaction-level investigation.

Landing Page & Network Selection

The Landing Page displays all networks available in your environment. Each network represents a distinct supply chain ecosystem generated for your mission. Each network card displays:

• Header image representing the network of interest
• Network name
• Last modified date
• Key statistics: total companies, facilities, transactions, and trade flows

Click any card to enter a network and open the Network Overview page.

Network Overview

When you open a network, the Overview provides a structured dashboard answering the fundamental question: what is the complete profile of this supply chain? The patterns you identify here — a geographic concentration, a dominant company, a risk cluster — become the questions you investigate in Flow View and beyond.

The Overview is organized around four analytical categories displayed as KPI cards at the top of the page: Supply Chain, Ownership, Exposures, and Product Analysis. Below these cards, a set of configurable visual panels provide at-a-glance breakdowns of the network’s most important characteristics.

Insights Sidebar

Click the Insights button in the upper left (below the navigation tabs) to open a sidebar with an AI-generated narrative overview of the network. This sidebar provides:

• A plain-language description of the supply chain and its structure
• A breakdown of each production phase with the specific HS codes and dollar values associated with it
• Key insights highlighting dominant patterns such as electronics concentration, vertical integration depth, or material diversity

KPI Categories

The four KPI cards at the top of the Overview each display a headline metric and a See More button that expands a detailed panel on that topic.

• Supply Chain — Headline: Total Companies, Total Facilities. Visuals include Top Companies by Country of Registration, Top Sectors by Company Count, Predicted Facility Types, Top Companies by Transaction Count and by Total Value (USD), Top Countries by Total Value and by Transaction Count, Top Companies by Connectivity, Top Production Phases by Company Count, and a Company Counts heatmap.
• Ownership — Headline: Foreign-Owned Companies, Total GUOs. Visuals include Top GUOs by Transaction Count, Top Countries of Registration by GUO, Top GUOs by Total Value (USD), and Top GUOs by Connectivity.
• Exposures — Headline: Exposed Companies, Exposure Types. Visuals include Top Exposure Types by Exposed Company Count, Top Countries of Registration by Exposed Company Count, and Top HS2 Codes by Exposed Company Count.
• Product Analysis — Headline: Total Transactions, Total Transaction Value (USD). Visuals include Top HS2, HS4 (with donut chart), and HS6 Codes by Total Value (USD), plus the same breakdowns by Transaction Count.

Tip: Use the Risk Assessment to find where foreign ownership intersects with risk exposure. If exposed entities concentrate in a specific country, filter to that country in Flow View to see which production phases are affected.

Toggling chart types: Each visual panel includes a small icon in the upper right corner. Clicking this icon toggles the chart between a bar chart view and a pie/donut chart view.

Reading percentages: The percentages shown next to each value (e.g., “912 (25%)”) represent that item’s share of the total across the network for that metric.

Customizing the Overview

Display Options (accessible from the top navigation bar) allow you to customize what appears on the Overview page:

• Select which visuals appear in the main overview slots: Add, remove, or rearrange the visual panels to focus on the metrics most relevant to your mission.
• Turn risk lists on or off: Display options control which risk and exposure datasets are active across the entire product (not just the Overview). Enabling or disabling a specific risk list here affects what appears in Exposures data, filter options, and entity profiles throughout the tool.

Tip: These display settings persist as you navigate between views, so configuring them once tailors the experience to your analytical priorities.

Flow View

Flow View, the core analytical space, visualizes a supply chain as an industrial production process organized left to right from upstream raw materials through downstream finished goods. You can see who controls each production stage, identify bottlenecks, and trace how disruptions might propagate.

Understanding the Layout

The visualization is organized into columns, each representing a production phase derived automatically using LLM analysis of the HS codes in the network:

1. The LLM processes all the HS codes to understand the industry.
2. The system derives a production sequence (e.g., raw material extraction → component manufacturing → sub-assembly → final assembly).
3. HS codes are assigned to phases using contextual LLM reasoning, not simple lookup.
4. Definitions for each phase and adjacent industries are refined iteratively.

Within each phase, entities appear in a treemap divided into HS code chapters (the default grouping, which you can change to alternatives such as Country or Company). Each leaf represents a company-owned facility, sized by transaction value or count.

You can show Adjacent industries below the primary phases by opening Display Options and clicking Show Adjacent Industries. These represent entities supplying goods necessary for production but not direct components (e.g., machinery for extraction is adjacent to, not part of, the primary product).

Upstream and downstream catch-all phases are not always present, but they may be added by the system if there are HS codes that fall outside the main derived production sequence. These phases capture entities shipping goods outside the production sequence; for example, raw materials might fall under an upstream column if they are not a primary focus of a specialized network.

How Entities Are Placed

• Entities are assigned to phases based on the goods they ship, identified by their HS code.
• For each transaction, the sender’s HS code determines placement.
• A company transacting multiple HS codes appears in multiple phases.
• Entities are rolled up to the company level: one appearance per company per phase.
• Placement is based on all HS codes transacted, not just top or highest-value codes.

Edge cases: Entities with unknown HS codes are placed before their receiver’s phase. Entities with out-of-scope goods land in upstream or downstream catch-all phases.

Inspecting Entities

Depending on size, each treemap leaf can display the company name, country (with flag), total USD value, and shipment count. Hovering over entities displays further details.

“Other Commodities”: Items below a visibility threshold are aggregated. The underlying data remains accessible through filters.

Selecting & Exploring Entities

Click an entity to select it (highlighted with a blue border). When you select an entity that appears in multiple phases, all instances of it are highlighted. Expand the Details panel to view more information about the selected entity.

Adjacent entities (direct trading partners) remain visible without a blue border, and non-connected entities fade to focus attention.

Multi-select is supported. When you click additional entities, they are added to your selection, and the Details panel and toolbar update for the group.

The Action Toolbar

When entities are selected, an action toolbar at the bottom of the screen helps you understand the network and work with the selection:

• Selected count vs. total: How many entities are selected out of the network
• Clear selection: Deselects all and resets the view
• View only selection: Hides everything except selected entities and connections
• Hide selection: Removes selected entities to focus on the remainder
• Add level upstream/downstream: Adds one tier of suppliers or customers to your selection
• Explore in Graph: Transfers selected entities to the Knowledge Graph for transaction-level investigation
• Zoom to Selection: Zooms in on selected entity (Map view only)

Tip: Use “add level upstream” iteratively to trace dependencies backwards. Select a company, expand upstream to see direct suppliers, and then expand again to see their suppliers to reveal multi-tier dependencies.

Interactive Network Summary

Expand the interactive Network Summary panel to find and highlight entities that correspond to production phases, top performers, or risks in the Flow View. Apply filters to automatically update the summary panel visuals and KPIs.

Filtering & Display Options

Filters and display options let you narrow analytical focus without losing broader context. Both are accessible from the top navigation.

Filters

Click Filters to access the following filter dimensions. Applied filters appear as tags below the nav bar.

• Companies: Find companies by name or ownership.
• Geography: Country of registration or GUO country.
• Products: HS codes at 2/4/6-digit levels with descriptions.
• Sector: Industry NAICS/NACE codes or BvD sectors.
• Risk & Exposure: Specific risk types or exposure categories.

Available filter options vary depending on the current view (e.g., Flow, Map, Global Trade).

Display Options

Click Display options to change how information is presented without removing data. Depending on the current view, you can change the following options:

• Color By: Country, Facility Type, GUO Country, Risk Exposure, Industry (NACE/NAICS), or BvD Sector.
• Group By: In the Flow View, changes how treemap leaves aggregate (HS code, Company, or Country).
• Size By: Total Exports (USD), Shipment Count, or Connectivity Score.

Available display options vary depending on the current view:

• Overview: Overview homepage visualization slots (select which visualizations to display in each slot); Exposures (choose which exposure types to show in overview and network dashboards).
• Flow: Color by, Size by, Group by (details outlined above).
• Map: Color by, Map fill (Choropleth).
• Global Trade: Color by.

Quick Assessment: For a foreign dependency scan, set Color By to “GUO Country” and Size By to “Total Exports (USD).” This instantly reveals which production phases are dominated by entities with foreign ultimate ownership, sized by trade volume.

Entity Details & Company Profile

When you identify an entity of interest, the Details Panel provides detailed context without leaving Flow View.

Opening the Profile

Select an entity, then click “Details” on the right. The panel shows the first selected entity with a total selected count. Click “See Company Profile” to open the full Altana profile in the Knowledge Graph.

Company Overview

• Total Transactions and Total Value within the network
• Country of Registration and GUO Country
• Sector and Industry (BvD, NACE/NAICS)
• Connectivity Score
• Percent of total network transactions and value

Company Details

• Goods Sent and Received (HS6 codes with descriptions)
• Top Trading Partners by shipment count (exports and imports)
• Production Phases the entity participates in
• Risk Exposures
• First and Last Known Activity dates

Facility Details

• Predicted Facility Roles
• Country
• Address
• Number of Transactions as Sender and as Receiver

Global Trade View (Chord Diagram)

The Global Trade View provides macro-level context by visualizing global trade flows using UN Comtrade data filtered to your network’s HS codes and timeframe. While the Flow View shows entity-level relationships, this view reveals which countries dominate trade in these goods and how flows interconnect globally.

This bridges Altana’s bottom-up transaction intelligence with top-down global trade statistics. Your network’s entities become a sampling of the macro data, enabling world-aggregate analysis anchored to specific examples.

Reading the Visualization

• Center = shipping activity (“ocean”). Edges = facilities in your network.
• Country bars are scaled to total global trade volume in relevant goods.
• Arrows show flow direction, pointing toward destination.
• Flow width represents the magnitude of connections between economic blocks.

Interpreting Flows

• Arrow reaching the edge bar: This country is a final destination.
• Connection not reaching the edge: This country is a pass-through point, not the final consumer.

Interacting

Hover over a perimeter facility to highlight the flow path for goods reaching that location. Facility interactions work the same as Flow View — select, view details, and transfer to the Knowledge Graph.

Map View

The interactive Map View provides a geospatial perspective. Facility locations reveal clustering patterns not always obvious in Flow View.

• Zoom and pan from a global view down to specific regions.
• Understand geographic clustering of suppliers and manufacturers.
• Perform country-level concentration analysis at a glance.
• Apply standard filters to focus on relevant data.

Tip: Map View is useful for co-location risk: are multiple critical suppliers in the same region, vulnerable to the same disruption?

Resilience Analysis

Resilience Analysis simulates how disruptions propagate through your supply chain. Based on published research on failure propagation in complex networks, it helps you identify vulnerabilities before they become crises.

How It Works

1. In any view, select specific nodes to disrupt, or let the system choose randomly.
2. Any selected nodes go “offline” for a specified time period.
3. Entities receiving from affected nodes have a probability of going offline based on remaining connections.
4. Recovery probability scales dynamically: losing 1 of 100 connections = high recovery; losing 1 of 2 = 50% chance of going offline.
5. The scenario runs multiple times and returns average outcomes.

Example: If you select 3 critical suppliers in Indonesia and run 20 simulations, the final reported Production Output will be the average predicted remaining production capacity of the network across those 20 runs.

Recovery Strategy Modes

• Any active edge: Recovery based solely on number of remaining connections. This pure topology model is the simplest and serves as a good starting point.
• All commodities provided: Considers specific goods (topology + HS Code). If only one edge provides a particular good, losing it eliminates recovery for that good — even with other connections for different products.
• All commodities provided (First two digits of HS Code): More flexible substitution (topology + two-digit HS Code). Losing HS 1234 but having HS 1256 (same two-digit parent 12) gives some recovery probability, reflecting real-world product substitution.
• Any new supplier: Searches the entire network for potential substitutions beyond existing connections. Powerful for identifying alternative suppliers you don’t currently have relationships with.

Running a Scenario

Select the entities you want to disrupt and then click Analytics. Configure the following parameters and recovery options:

• Selection Initially Damaged: Fill the check box to disrupt the selected nodes or clear it to let the system choose randomly.
• Damaged Nodes (%): Random subset (only if no manual selection).
• Edge-Switch Probability: Recovery speed. Higher = faster.
• Number of Runs: Simulation repeats for averaging. 10–20 for quick estimates; 50+ for reports.
• Recovery Time (Weeks): Duration damaged nodes stay offline.
• Node Recovery Strategy: One of the four modes above.

The simulation typically takes a few minutes for average-sized networks with default settings, but runtime can be longer. Calculation time depends on overall network size, number of iterations, and number of selected entities.

Interpreting Results

When the scenario is complete, the results display automatically. You can navigate away from the page while the simulation is running; results can be retrieved in the Analytics results tab.

• Overall Resilience: Ability to recover over time. Higher is better.
• Overall Robustness: Ability to withstand initial disruption. Higher is better.
• Production Output: Estimated remaining production capacity. The most tangible metric for decision-makers.

Key Insight: Resilience Analysis is most powerful as a comparative tool. Run baseline scenarios, then re-run after removing high-risk nodes. The difference quantifies the risk each entity poses to the network.

Chat (AI Assistant)

Chat is the natural language interface, powered by Anthropic’s Claude through Altana’s enterprise license. Click “Chat” in the upper right corner to open the Chat panel.

What Works Well

Resilience Scenarios (core strength):

• “What happens if facilities in China go offline?”
• “Run a disruption scenario on [Company Name]”
• “What happens if upstream entities in Singapore go offline for 6 weeks?”

Filtering & Selection:

• “Show me only Chinese suppliers”
• “Filter to companies with sanctions exposure”
• “Display facilities in the components phase”

Network Characterization:

• “What is this supply chain about?”
• “Who are the key players?”
• “What are the most vulnerable parts?”
• “What are the data gaps?”

Geographic & Trade Analysis:

• “What entities are upstream in Chile?”
• “If there is a problem with the China-US lane, what are the biggest flows?”
• “Who are the top customers of [Company] within 3 steps downstream?”

Tips for Best Results

• Be specific: “Chinese suppliers in the components phase” not “Show me China”
• Specify metrics: “Top 5 by transaction value” not “Top 5”
• Reference specific production phases
• Use entity names exactly as they appear
• Break complex questions into simpler queries
• Always verify filter application visually

Current Limitations

These query types are not yet reliably supported:

• Entity recognition from selections: “Tell me about these selected suppliers” does not yet recognize visual selections.
• Structured ranked outputs: “Top 5 alternative suppliers” cannot yet produce structured lists.
• Complex multi-criteria logic: Combining multiple filter dimensions in one query is unreliable. Apply incrementally.
• Temporal analysis: Time-series questions are not yet supported.

Workaround: For analytics Chat can’t handle yet, use Flow View filters, Display Options, and the Network Summary panel directly. These visual tools are reliable for all query types.

Security & Privacy

Queries are processed through Anthropic Claude via Altana’s enterprise license. No queries are stored, retained, or used for model training. Server console logs exist for debugging only and are not attributable to specific users.

Knowledge Graph Integration

Network Analysis provides the macro view. The Altana Knowledge Graph provides the micro detail: transaction histories, ownership structures, detailed profiles, and entity-level investigations. They are connected by design.

Transferring Entities

1. Select 1–30 entities in Flow View.
2. Click “Explore in Knowledge Graph” in the action toolbar.
3. The Knowledge Graph opens with selected entities pre-loaded.
4. Expand supply chains, view transaction histories, investigate ownership, conduct compliance analysis.

Performance: Transfer ~10 or fewer entities at a time for best results. Use filters to narrow to the highest-priority entities first.

When to Use

• View full transaction history between two entities
• Investigate beneficial ownership chains beyond the GUO
• Access the complete Altana Company Profile
• Expand the supply chain beyond the pre-built network boundaries
• Collect evidence for enforcement actions or compliance determinations

Platform Walkthrough Video

Watch the full Network Analysis platform walkthrough: