Build market intelligence workflows with Vyfin

Quick start

Use this playbook to recreate the equity exploration we show in product tours. Every action reflects the Nasdaq 100 disclosure set that feeds Vyfin’s metrics.

1. Calibrate descriptors and baselines

   Load the equity canvas and apply the Descriptor overlay. Add Dividend Yield, Market Cap, and Revenue Growth, then tune thresholds to tier companies into comparable cohorts. The workspace normalizes shorthand values such as 100M or 2B, so you can focus on analysis instead of formatting.

2. Balance dividend yield and revenue

   Set your preferred ranges for each filter to surface balanced income and growth candidates. Threshold adjustments immediately refresh the canvas and preserve undo history for transparent iteration.

3. Layer Beta and investigate single names

   Add Beta with ranges that match your risk appetite, then click a highlighted node to open the analysis drawer. Use the 📈 Chart tab to review price action, volume, and comparative benchmarks, and pivot to the 🧾 Financials tab to inspect the filing-backed metrics that explain each inflection. Every figure pulls from the latest Nasdaq 100 10-Q and 10-K disclosures, so the chart and the statements stay in lockstep.

4. Use multi-select to focus the thesis

   Switch to the lasso tool and drag a marquee across the standouts. The floating panel offers Subnetwork and Add to Watchlist actions so you can isolate narratives for different teams. Promote a selection into Subnetwork to spin up a fresh canvas with its own physics, notes, and comparison metrics; rename the slice, experiment with alternate encodings, and publish screenshots without disturbing the main view. Save the same cohort into your company list to revisit it later, tag the scenario for sales, product, or executive reviews, and keep iterative snapshots aligned with the underlying filters. Hold Ctrl (⌘ on macOS) to add or remove individual names without breaking flow.

5. Tune layout, scale, and color encodings

   Dial in physics from the layout tool with repulsion, link distance, and collision controls. Activate Size for Total Revenue and Color for Dividend Yield, Revenue Growth, and Debt To Equity to surface balance-sheet posture at a glance. Layer labels, hulls, outliers, communities, and opacity to reinforce the story.

6. Capture the story for sharing

   Review the Tables, Stats, and Logs tabs to document the quantitative rationale. Save the configuration, export SVG or PNG assets, and rely on Back, Reset, and undo history to iterate without losing your preferred layout.

Product overview

Vyfin’s second-generation canvas blends force-directed mapping, descriptor tiers, and community detection to translate fundamentals into spatial intuition. Analysts can retrace every move: company nodes plot as circles, descriptor buckets surface as distinct shapes, and live metrics scroll overhead to anchor the narrative.

The latest release shipped the analytics-heavy 📊 Stats tab, revamped watchlist import/export flows, RGB color mapping, and premium indicators tuned to the Nasdaq 100 disclosure pipeline. Interactions stay synchronized across the workspace so product tours and documentation present the same state.

• Descriptor intelligence: Communities, hulls, and descriptor palettes surface the structural forces behind each research storyline.
• Live fundamentals: The metric banner and tables stream the latest values derived from Nasdaq 100 10-Q and 10-K filings without manual refreshes.
• Guided analytics: Quick links to TradingView, financial statements, and watchlists keep research reproducible.

Architecture runbook

Vyfin pairs a modular browser workspace with a secure service layer. The front end manages tool state, event orchestration, and rendering so analysts can pivot between configurations without reloads.

The service layer prepares JSON payloads that join equity fundamentals, descriptor metadata, and analytics derived from Nasdaq 100 filings. Requests are authenticated, logged, and optimized for responsiveness across research teams.

Subscription status, watchlists, and collaboration cues synchronize in real time so upgrades, saves, and shared canvases appear instantly for every collaborator.

Filters & data pipeline

The filter catalog organizes metrics by theme so teams can mix dividend, growth, quality, and risk perspectives. Free plans surface Dividend Yield, Market Cap, and Beta, while premium subscriptions unlock the full financial stack.

When you confirm a metric, the platform pulls the latest thresholds from Nasdaq 100 10-Q and 10-K filings and applies them to the canvas. Every adjustment updates company nodes, descriptor groupings, and tool payloads while maintaining debounced inputs, undo checkpoints, and watchlist warnings for transparency.

Tool reference

Network interactions

Clicking a node recenters the legend, updates the metric banner, and spawns external research affordances (TradingView, fundamentals) without leaving the canvas. Dragging nodes lets you inspect local structures; D3 temporarily fixes the dragged node so manual placements hold during analysis.

Undo/redo (Ctrl/⌘ + Z and Ctrl + Y/⌘ + Shift + Z) keeps experiments reversible. The ⬅️ Back button restores the previous filter + layout snapshot after Subnetwork drills, while ♻️ Reset returns the full network to fit. All interactions—including watchlist updates—stream to 📜 Logs for post-hoc auditing.

Colors & legends

Descriptor nodes follow the small/medium/large palette (green/orange/red) and shapes (square/triangle/diamond). Without RGB or community overrides, company nodes blend their descriptor colors to reveal factor exposure at a glance.

Activating 🎨 Color rewrites fills according to your metric assignments. The legend and metric banner capture the mapping, while 👥 Communities swaps to the modularity palette we use for clustering. Hulls, outlier halos, and descriptor edges all stay in sync so exported assets translate cleanly to presentations.

Analytics tabs

How the Stats tab builds context

📊 Stats projects the descriptor-company bipartite graph into a company-only view before ranking scores. Each descriptor is treated as a cohort, so shared exposures determine how tightly two companies relate within the current filters.

The projection pipeline standardizes each descriptor’s company weights between zero and one, then averages the paired values every time two companies appear together. Those averaged weights accumulate into a symmetric adjacency matrix, and any edge without meaningful overlap is discarded. The resulting network emphasizes disclosure-based similarity instead of raw filing counts.

Centrality scores derive directly from that matrix:

• Genericness (degree) counts how many neighbors carry a positive weight. Because the network is undirected, every qualifying connection increments the score once, making it a pure measure of breadth.
• Mainstream (eigenvector) runs a power-iteration loop: start every company at one, replace each score with the weighted sum of its neighbors, normalize by the vector length, and repeat until the numbers settle. Companies tethered to other influential names rise to the top.
• Popularity (closeness) applies a weighted shortest-path search from every origin, interpreting distance as the reciprocal of edge weight. The score takes the count of reachable peers and divides it by the total distance, so densely connected companies earn higher values.
• Niche Bridge (betweenness) adapts the Brandes algorithm for weighted graphs. It traces all shortest paths discovered with the same distance rule and assigns credit whenever a company sits between two others. The undirected adjustment halves the total so shared routes are not double-counted.

Use the count and order controls to decide whether you want the top or bottom performers, then interpret the charts with the helper heuristics surfaced in the app:

• Genericness (degree) captures how many close peers a ticker keeps under the current filters; low values spotlight one-off structures.
• Mainstream (eigenvector) rewards names connected to influential peers, while Popularity (closeness) reveals how quickly you can relate a company to the broader cohort.
• Niche Bridge (betweenness) climbs when a company links otherwise disconnected descriptor themes—prime territory for differentiated theses.

The helper combos translate those signals into narratives: Benchmark Poster Child (High Genericness + High Mainstream, Low Niche Bridge) surfaces benchmark staples, Contrarian Crossover (High Niche Bridge with Mid/Low Mainstream) spotlights mispriced bridges, and Overlooked Bridge (High Niche Bridge + Low Genericness) flags unique structures worth validating.

When a pattern stands out, jump back to the canvas with lasso selections or watchlists and verify the same story visually before exporting.

Reading the Tables view

📋 Tables rebuilds itself from the live workspace state. The renderer starts with Ticker and Cluster, then appends every active filter metric plus the metrics powering RGB color, size, and opacity. Values stream from the filings-backed dataset so your spreadsheet or export matches the on-screen encoding.

Work through the table deliberately to keep the quantitative story airtight:

• Scan the Cluster column alongside Stats rankings to see which outlier groups drive high or low structural scores.
• Blank cells highlight metrics missing from the latest filings, prompting a re-check before you commit a number to stakeholders.
• Copy snapshots after major filter tweaks to create an audit trail that lines up with the 📜 Logs tab.

Keep Tables open while preparing exports so every annotation and callout references a number you can cite instantly.

Watchlists & sharing

The Company Watchlist table tracks 📥 additions with chart, fundamentals, and delete affordances. Rename each list to match the narrative you are building, drag rows to prioritize coverage, and jump straight into the 📈 Chart or 🧾 Financials tabs for any saved company without hunting through the canvas. Clear Watchlist resets the slate before your next case study.

The Filter Watchlist stores saved configurations, applies tool settings on load, and highlights premium-only gaps for free users. Use descriptive labels to flag whether a setup aligns with income, growth, or thematic hunts, and rely on the preview chips to confirm which metrics, colors, and physics settings will load.

🔁 copies a JSON payload for sharing, while Import Watchlist validates pasted JSON and surfaces warnings when a configuration references locked metrics. Pair Subnetwork exports with the corresponding saved lists so collaborators understand which focused canvases to load, which companies to prioritize, and how the scenario evolved across iterations.

Export & history

Use 💾 SVG for vector exports and 🖼 PNG for ready-to-publish images. The client serializes the SVG, renders it to canvas, and triggers a download—perfect for embedding in stakeholder summaries.

History state captures every filter, layout, and tool change so you can iterate freely. ⬅️ Back restores pre-Subnetwork views, ♻️ Reset recenters the canvas, and the undo/redo stack keeps experiments safe.

Resource hub