Awesome AI Market Maps

Aggregates 400+ AI market maps from 50+ sources (Tier 1 VCs, specialized investors, analysts) into a unified README.md single-source-of-truth using a two-dimensional taxonomy (temporal quarters/months × thematic AI domains). Implements hierarchical markdown structure with level-2 headers for quarters and level-3 headers for months, enabling deterministic parsing by downstream automation pipelines. The architecture enforces unidirectional data flow where README.md is the canonical source, preventing synchronization conflicts across derivative outputs (RSS, CSV, external platforms).

Transforms README.md markdown structure into valid RSS/XML feed via GitHub Actions workflow (re-build-rss.yml) that runs on push events. The generate_rss.py script parses markdown hierarchically starting from the '## ▦ MARKET MAPS ▦' delimiter, extracts market map entries with metadata (title, source, date, URL), sanitizes text for XML compatibility, and generates timestamped RSS entries. Implements real-time syndication with near-zero latency between README.md updates and feed availability, enabling subscribers to receive new market maps via RSS readers without polling the repository.

Integrates with external platforms (Twitter, LinkedIn, Slack) to republish market map updates beyond the GitHub repository. Market map additions can be automatically or manually cross-posted to these platforms, extending reach to audiences who don't follow the GitHub repository directly. Integration points include Twitter API for tweet posting, LinkedIn API for article sharing, and Slack webhooks for channel notifications. This capability enables the market map collection to function as a content distribution hub, with GitHub as the source of truth and external platforms as distribution channels. Cross-posting can be triggered manually by the maintainer or automated via GitHub Actions workflows.

Enables researchers and analysts to discover relevant market maps for specific AI domains, time periods, or source organizations through browsing, filtering, and searching capabilities. Users can navigate the hierarchical README.md structure to find maps by quarter/month or domain, use CSV export to filter programmatically, or subscribe to RSS feed for specific categories. The repository also serves as a research artifact itself, enabling meta-analysis of market map creation patterns (e.g., 'which domains have the most maps?', 'how has VC focus shifted over time?'). This capability transforms the collection from a passive list into an active research tool for understanding AI market evolution.

Exports aggregated market map metadata into a structured CSV dataset (ai_market_maps.csv) with columns for date, source organization, market map title, AI domain category, and direct URL link. The export is manually maintained with documented lag (typically bimonthly refresh cycle), allowing downstream tools (Pandas, Excel, Tableau, SQL databases) to ingest market map data for analysis, filtering, and visualization. Provides a machine-readable alternative to markdown for users who need tabular data structures, enabling programmatic access without parsing markdown syntax.

Distributes aggregated market map data across three output formats (Markdown README, RSS feed, CSV export) with intentionally different update cadences: README.md updates on manual edits (immediate), RSS regenerates on every push (near-real-time), and CSV refreshes bimonthly (batch). This tiered freshness strategy allows different consumer personas to choose their preferred trade-off between recency and stability. The architecture maintains unidirectional data flow from README.md as single source of truth, preventing synchronization conflicts while enabling each format to optimize for its use case (human browsing, feed subscription, data analysis).

Implements a fixed taxonomy of AI domain categories (agents, RAG, code generation, image generation, etc.) used to classify and organize market maps within the README.md structure. Market maps are grouped by both temporal dimension (quarters/months) and thematic dimension (AI domain), enabling discovery along either axis. The taxonomy is curated by the repository maintainer and applied consistently across all 400+ market maps, allowing users to filter by domain (e.g., 'show me all agent-related market maps') or track how investor attention shifts within specific domains over time.

Organizes market maps along a temporal dimension using hierarchical markdown headers: level-2 headers for quarters (e.g., '## AI Market Maps - Q1 2026') and level-3 headers for months (e.g., '### January 2026'). This structure enables users to browse market maps by publication date, track how market maps evolve within specific time periods, and identify temporal trends (e.g., 'which domains had the most maps in Q4 2025?'). The temporal hierarchy is deterministically parseable by automation scripts, allowing RSS generation and CSV export to preserve publication dates and enable time-based filtering.

Applies manual curation to maintain Awesome List standards, where the repository maintainer (Joy Larkin) reviews candidate market maps for inclusion based on subjective 'taste' criteria. The curation process filters out low-quality, duplicative, or off-topic maps before adding them to README.md, ensuring the collection remains focused and high-signal. Curation decisions are made asynchronously via GitHub issues, pull requests, and direct contributions, with the maintainer as final arbiter of inclusion. This human-in-the-loop approach trades scalability for quality, keeping the collection curated rather than exhaustive.

Enables community contributions via GitHub pull requests and issues, allowing users to propose new market maps for inclusion in the collection. Contributors submit market map URLs with metadata (title, source, date, category) following documented guidelines, which are then reviewed by the maintainer for quality and relevance. The workflow is asynchronous and GitHub-native, requiring no external tools or registration beyond a GitHub account. Accepted contributions are merged into README.md, while rejected submissions receive optional feedback. This distributed contribution model allows the repository to scale beyond the maintainer's personal discovery while maintaining quality control.

Extracts and normalizes metadata from market map entries in README.md, including title, source organization, publication date, AI domain category, and direct URL link. The extraction process is deterministic, parsing markdown link syntax and hierarchical headers to identify metadata fields. Normalization includes date standardization (converting various date formats to YYYY-MM-DD), source name deduplication (e.g., 'a16z' vs 'Andreessen Horowitz'), and URL validation. This normalized metadata feeds downstream systems (RSS generation, CSV export, external platform integrations), enabling consistent data representation across all output formats.

Integrates Git commit history to infer or validate publication dates for market maps when explicit dates are unavailable or uncertain. The system can query Git metadata (commit timestamps, author information) to determine when a market map entry was added to the repository, providing a fallback date source. This capability enables temporal analysis even when source market maps lack explicit publication dates, and allows the system to track when maps were discovered/added relative to their actual publication. Git history integration is used by the RSS generation pipeline to populate pubDate fields and by temporal analysis tools to understand discovery lag.