Network & Relationship Intelligence

See the summary view → for a quick overview of all Network & Relationship Intelligence capabilities.

16 capabilities powered by a knowledge graph with 200K+ entities and graph algorithms that analyze the structure of legislative relationships - not just individual data points, but the patterns of connections between them. All capabilities are live and available.

What makes this different: Traditional platforms show you data. Network intelligence shows you relationships - who influences whom, which coalitions are forming, where the power centers are, and how entities connect across the political ecosystem. These analytics use PageRank, community detection, Jaccard similarity, and shortest-path algorithms running on live legislative data.

Member Universe

Comprehensive single-query profile of any member of Congress pulling together their complete legislative footprint: sponsored and cosponsored bills, committee assignments, media coverage, organizational stances (support/oppose from press releases), PAC contributions received, network allies, and graph-derived influence metrics. Instead of running six separate searches, one query returns the full picture.

The member universe is powered by a graph traversal that follows Member → Bill, Member → Committee, Member → News, Member → PAC, and Member → Ally relationships in the knowledge graph, assembling a multi-section profile with data from every source Apogee tracks.

How it works: A single graph query traverses all relationship types from a Member node in the knowledge graph, returning structured sections for legislation, committees, media, funding, allies, and stances.

Data: Knowledge graph combining Congress.gov member profiles, bill sponsorship records, committee assignments, news entity mentions from 40+ outlets, PAC contribution records (FEC), and press release stances.

Tell me everything about Nancy Pelosi's legislative network Try it → Give me a full profile of Senator Warnock - bills, committees, allies, media Try it →

Member Search & Lookup

Search and filter members of Congress by state, chamber, party, congressional district, or committee assignment. Find the right member quickly with basic profile information including name, party, state, district, chamber, and current committee memberships.

This is a foundational lookup capability that other tools build on - cosponsor analysis, influence ranking, and alliance detection all start with identifying the right members. Supports partial name matching and handles common name variations.

Data: Congress.gov member profiles for the current Congress, including party affiliation, state, district, chamber, and committee assignment data. Updated as committee assignments change.

Who represents California's 12th district? Try it → List all members of the Senate Armed Services Committee Try it → Show me Democratic senators from swing states Try it →

Co-sponsorship Alliance Analysis

Identify a member's closest legislative allies based on how frequently they cosponsor the same bills. Two complementary algorithms reveal different dimensions of collaboration: frequency-based analysis counts raw shared bill cosponsorship (who works together most often), while Jaccard similarity controls for prolific cosponsors by measuring profile overlap as a proportion of total activity (who has the most similar legislative profile).

The distinction matters. A member who cosponsors 200 bills will share many cosponsors with everyone by volume. Jaccard similarity normalizes for this, revealing the members whose sponsorship choices are genuinely aligned rather than just prolific. For government affairs teams identifying potential allies for a legislative campaign, Jaccard similarity is the more reliable signal.

How it works: The Jaccard algorithm runs on a graph projection of the cosponsorship network. For each member pair, it computes |shared bills| / |total unique bills between them|. Frequency mode uses a simpler shared-count query.

Data: Congress.gov cosponsorship records for the current Congress, modeled as cosponsorship relationships in the knowledge graph.

Who are Senator Collins' closest legislative allies? Try it → Find bipartisan collaborators for the chair of the Energy Committee Try it → Which Democrats most frequently cosponsor bills with Republican members? Try it →

Entity Connection Discovery

Shortest-path traversal between any two entities in the knowledge graph - members, bills, committees, PACs, organizations, or lobbying firms. Reveals how two seemingly unrelated entities connect through the political ecosystem, showing every intermediate node and relationship type along the path.

This is the "six degrees of separation" for politics. A PAC might connect to a bill through a member who received contributions and sits on the committee of referral. An organization might connect to a piece of legislation through its lobbyists, the members those lobbyists previously served under, and those members' committee assignments. These connection chains are invisible without graph traversal.

How it works: Dijkstra's shortest-path algorithm traverses all relationship types in the knowledge graph - sponsorship, cosponsorship, contributions, lobbying, committee membership, and media mentions. Returns the complete path with node labels and relationship types at each step.

Data: Full knowledge graph with 200K+ entities and millions of relationships spanning members, bills, committees, PACs, organizations, lobbying firms, and news articles.

How are Koch Industries connected to the Farm Bill? Try it → What's the shortest path between Senator Warren and the Chamber of Commerce? Try it → Show me the connection between the Lockheed Martin PAC and the National Defense Authorization Act Try it →

Topic Connection Discovery

Find how an entity connects to an entire policy area or topic - not just a single entity. While entity connection discovery requires exact node IDs for both endpoints, topic connection discovery takes a named entity and topic keywords, then finds the shortest paths to any bills, hearings, or committees matching that topic across the knowledge graph.

This is the tool for questions like "how is the Chamber of Commerce connected to AI regulation?" where the target isn't a specific bill or committee but a broad policy concept spanning multiple nodes. Instead of requiring 15+ separate tool calls to manually stitch together connections, a single graph query finds all relevant paths in sub-second time.

How it works: The source entity is resolved by name (with optional type hint for organizations, members, or PACs). Target nodes are matched across Bill titles, Hearing titles, and Committee names using case-insensitive keyword matching. Neo4j's shortestPath() algorithm then finds the shortest path from the source to each matching target, returning up to 10 paths ranked by hop count.

Data: Full knowledge graph with 200K+ entities. Source entities matched against Organization (canonical names + aliases), Member (names + aliases), and PAC nodes. Target nodes matched across Bill (title + allTitles), Hearing (title), and Committee (name) labels.

How is the Chamber of Commerce connected to AI regulation? Try it → What links Koch Industries to climate policy? Try it → Show me how the NRA connects to gun control legislation and hearings Try it →

Voting Bloc Identification

Community detection algorithms reveal the actual voting coalitions forming in Congress - which may not align neatly with party labels. Members who vote together on 10+ shared roll-call votes with greater than 50% agreement are grouped into blocs using the Louvain algorithm, producing a map of voting communities that reflects real legislative behavior rather than assumed party loyalty.

The most interesting findings are often at the edges: moderate members who cluster with the opposing party on specific issue areas, or intra-party factions that vote differently from their leadership on key legislation. For government affairs professionals, understanding these blocs helps identify which members might break from their party on a given issue.

How it works: A bipartite graph of voting relationships is projected for analysis. The Louvain community detection algorithm groups members by voting similarity, producing communities with modularity scores. Results include bloc membership, intra-bloc agreement rates, and cross-bloc relationships.

Data: Roll-call vote records from Congress.gov mapped as voting relationships in the knowledge graph. Updated as new roll-call votes are recorded.

What voting bloc does Senator Manchin belong to? Try it → Show me the current voting coalitions in the House Try it → Which voting blocs cross party lines? Try it →

Member Influence Ranking

Rank members of Congress by network influence using two complementary graph centrality algorithms: PageRank measures collaborative influence (who do other members work with most, recursively - like Google's original algorithm applied to legislative relationships), while Betweenness Centrality measures bridge position (who connects otherwise-separated groups in the collaboration network).

A member with high PageRank is well-connected to other well-connected members - the legislative hubs. A member with high Betweenness Centrality sits on the shortest paths between different groups - the brokers and dealmakers. These two measures together identify both the most connected and the most strategically positioned members in Congress.

How it works: PageRank and Betweenness Centrality algorithms run on a projection of the cosponsorship collaboration network. Results are normalized and can be filtered by chamber, party, or committee.

Data: Cosponsorship collaboration network derived from Congress.gov bill data, modeled as weighted Member → Member relationships based on shared bill cosponsorship frequency.

Who are the most influential members in the Senate by network position? Try it → Which House members are the most important bridge-builders? Try it → Rank members of the Judiciary Committee by legislative influence Try it →

Legislative Broker Detection

Finds members who bridge different voting blocs - the cross-coalition connectors who build consensus across otherwise-separated groups. While member influence ranking identifies overall network position, broker detection specifically targets members who span community boundaries, making them uniquely positioned to assemble bipartisan or cross-faction coalitions.

These are the members that government affairs teams should prioritize when building support for legislation that needs cross-party buy-in. A broker might not be the most influential member overall, but they have relationships in multiple camps - making them the linchpin for coalition-building.

How it works: Combines community detection (which voting bloc each member belongs to) with bridge scoring (how many cross-bloc relationships a member has). Members with high bridge scores relative to their bloc size are identified as brokers.

Data: Roll-call vote records and cosponsorship data modeled in the knowledge graph, with community membership computed by the Louvain algorithm.

Who are the bipartisan bridge builders in the Senate? Try it → Which members connect the progressive and moderate Democratic blocs? Try it → Find the legislative brokers on the Armed Services Committee Try it →

Unusual Alliance Detection

Surfaces pairs of members who cooperate despite being in different parties or different voting blocs - identifying unexpected collaborations that may signal emerging consensus on an issue, cross-partisan deal-making, or shared constituent interests that transcend party lines.

When two members from opposing parties or different voting communities consistently cosponsor the same legislation, it's a signal worth investigating. These unusual alliances often form around geographic proximity (neighboring districts), shared constituent industries, personal relationships, or emerging policy consensus that hasn't yet registered as a party-wide position.

How it works: Pair analysis on the cosponsorship graph, filtering for member pairs who belong to different voting communities (as detected by Louvain) but share significant cosponsorship overlap. Ranked by cooperation frequency relative to expected baseline.

Data: Cosponsorship records and voting community memberships from the knowledge graph, combined to identify cross-community collaboration patterns.

What unusual bipartisan alliances have formed this Congress? Try it → Which cross-party pairs cosponsor the most legislation together? Try it → Find surprising collaborations on tech policy bills Try it →

Committee Power Mapping

Weighted influence analysis per committee - identifies which members hold the most network power within a specific committee, going beyond formal titles (chair, ranking member) to reveal the actual power structure based on legislative activity, cross-committee connections, and collaboration centrality.

A committee member who sponsors heavily-cosponsored legislation, connects to members on related committees, and bridges different factions within the committee may wield more practical influence than a senior member with a higher formal rank. Committee power mapping surfaces these dynamics, which are critical for lobbyists, advocates, and policy strategists deciding which committee members to engage.

How it works: Centrality algorithms (PageRank and Betweenness) run on a subgraph of the cosponsorship network filtered to a specific committee's membership. Results are weighted by bill activity within the committee's jurisdiction.

Data: Committee membership rosters from Congress.gov, combined with cosponsorship collaboration data and bill-committee referral relationships in the knowledge graph.

Who are the most influential members on the Senate Finance Committee? Try it → Map the power structure of the House Energy and Commerce Committee Try it → Which committee members have the most cross-committee influence? Try it →

Cosponsor Prediction

Forecast likely cosponsors for any bill based on historical collaboration patterns and sponsorship profile similarity. Returns ranked predictions with overlap scores - identifying members who are statistically likely to support a bill based on how closely their past cosponsorship choices match the bill's existing sponsor coalition.

This is a prospective capability: given a bill's current sponsors and cosponsors, who else is most likely to sign on? For government affairs teams running whip operations or advocacy campaigns, this narrows the field from 535 members to a ranked list of realistic targets based on demonstrated legislative behavior, not just party affiliation or committee assignment.

How it works: Jaccard similarity analysis compares the bill's current cosponsor set against each member's historical cosponsorship profile. Members with the highest overlap score (most similar sponsorship patterns to the existing coalition) are ranked as most likely future cosponsors.

Data: Congress.gov cosponsorship records for the current and recent Congresses, modeled as Member → Bill relationships in the knowledge graph.

Who is most likely to cosponsor semiconductor manufacturing legislation in the current Congress? Try it → Predict the coalition for the Clean Energy Innovation Act Try it → Which Republicans are most likely to support the bipartisan infrastructure proposal? Try it →

Similar Bills Discovery

Find legislation with similar sponsorship profiles using Jaccard similarity on the cosponsorship graph. Bills that share the same sponsors and cosponsors score higher - surfacing related legislation that may not share keywords but attracts the same political coalition, revealing ideological alignment that text-based search misses entirely.

Two bills about completely different topics (say, a telecommunications reform bill and a rural broadband bill) might share 80% of the same cosponsors - suggesting they appeal to the same political coalition and may face similar legislative dynamics. For policy strategists, discovering these sponsorship-profile similarities reveals legislative patterns invisible to keyword search.

How it works: Jaccard similarity analysis computes the overlap between each bill pair's cosponsor sets. Bills are ranked by similarity score, with results including the specific shared cosponsors and unique-to-each-bill cosponsors.

Data: Congress.gov cosponsorship data for all bills in the current Congress, modeled as cosponsorship relationships.

Find bills with similar support to the Infrastructure Investment Act Try it → What other legislation attracts the same cosponsors as the Future of Artificial Intelligence Innovation Act? Try it →

Media Mention Tracking

Graph-based entity mention search across news coverage - tracking how media attention is distributed across members, committees, organizations, and bills using entity-linked articles from the knowledge graph rather than simple keyword matching.

Unlike keyword search (which returns articles containing a name string), entity-linked search resolves mentions to canonical entities. Searching for "Senator Warren" returns articles that mention Elizabeth Warren by any name variation, including articles that reference her by title, committee role, or contextual description. This produces more complete and accurate media tracking.

How it works: Graph traversal from a Member, Committee, or Organization node through media mention relationships to articles. Results include article metadata, publication date, source outlet, and co-mentioned entities.

Data: 36,000+ articles from 50+ policy news sources with AI-extracted entity mentions linked to knowledge graph nodes. New articles ingested via RSS feeds multiple times daily.

Which members are getting the most media coverage on AI policy? Try it → Track media mentions of the Senate Judiciary Committee this month Try it →

Organization Influence Network

Unified view of an organization's complete political footprint in a single query: news coverage mentioning the organization, press release stances (support/oppose on specific bills), entity relationships (connected members, committees, and bills), lobbying filings and expenditures, and PAC contribution patterns. Answers "tell me everything about Organization X's political activity" without requiring separate searches across six different databases.

This eliminates the most common multi-query pattern in policy research. Understanding an organization's full political profile previously required cross-referencing FEC filings, SOPR lobbying disclosures, news databases, and press release archives manually. Apogee's organization universe traverses all these relationships from a single entity node in the knowledge graph.

How it works: A single graph query traverses all relationship types from an Organization node - news mentions, stances, lobbying filings, PAC connections, and member relationships. Supports lookup by name with section toggle flags.

Data: Knowledge graph combining FEC PAC data, SOPR lobbying disclosures, press release entity extraction with stance analysis, and news coverage from 40+ outlets - all linked through entity resolution to canonical organization nodes.

Map the political influence network of the Chamber of Commerce Try it → Show me everything about Lockheed Martin's political activity Try it → What's AARP's full footprint - lobbying, stances, mentions, connections? Try it →

Policy Community Detection

Community detection across the cosponsorship network filtered by policy area - identifying which members of Congress form coherent legislative communities around specific policy domains. While voting bloc identification groups members by voting patterns, policy community detection uses the Louvain algorithm on bill-scoped collaboration networks to reveal the ecosystems of legislators that form around healthcare, defense, technology, or any other policy area.

Understanding which members form a coherent policy community is essential for government affairs teams planning engagement on an issue. A member deeply embedded in the healthcare policy community - even without a formal committee assignment - may be a more effective advocate than a committee member with no collaboration history on the topic.

How it works: The Louvain community detection algorithm runs on a graph projection of cosponsorship relationships filtered to a specific policy area. Members are grouped into communities based on shared bill activity within that domain. Results include community membership, member profiles, and policy area context.

Data: Congress.gov cosponsorship records and bill policy area classifications, modeled as relationships in the knowledge graph. Graph analytics recomputed every 6 hours.

Who are the key players in the healthcare policy community? Try it → Map the ecosystem of members focused on AI and technology policy Try it → Which members form the core legislative community around defense spending? Try it →

Rising Influence Detection

Identifies members whose legislative influence is growing faster than expected by comparing PageRank centrality scores across consecutive Congresses, adjusted for seniority. Members who are newer to Congress but climbing the influence rankings faster than their tenure would predict are flagged as rising power brokers - the legislators building outsized networks through strategic cosponsorship and cross-party collaboration.

Detecting rising influence early is valuable for government affairs teams that want to build relationships before a member becomes a gatekeeper. A backbencher whose centrality is climbing quickly is often one committee assignment or leadership election away from becoming a key player. The seniority adjustment ensures freshmen members who are genuinely outperforming are distinguished from members who are simply accumulating connections over time.

How it works: Congress-scoped graph projections of the cosponsorship network are created for two consecutive Congresses (e.g., 118th and 119th). PageRank is streamed on each projection to compute per-member influence scores, then a composite score is calculated from three signals: PageRank delta (50%), cosponsorship velocity (30%), and seniority factor (20%). The seniority adjustment uses 1 / sqrt(yearsInOffice) so freshmen get the highest multiplier. Results can be filtered by chamber, party, or policy area.

Data: Congress.gov cosponsorship records across multiple Congresses and member term history, modeled in the knowledge graph. Requires GDS projections computed on demand.

Which freshmen members are building the most legislative influence? Try it → Who are the rising power brokers in the Senate? Try it → Which members have gained the most influence in healthcare policy compared to last Congress? Try it → Show me Democrats whose legislative network is growing fastest Try it →

Money & Influence Intelligence - PAC contributions, lobbying, funding anomalies
Legislative Intelligence - Bill search, cosponsor analysis, momentum scoring
Strategic Intelligence - Coalition building and competitive landscape mapping

Ready to try these capabilities? Get started with Apogee or try the chat client.