Category · AI Citation Index
AI Infrastructure
Databricks leads 90-brand field at 80% shortlist rate; vector DBs and experiment trackers fragment mid-tier; H2O collapses 63 points.
660 discovery queries · 883 head-to-heads · refreshed Sep 25, 2026
Discovery stage
The shortlist
Across 660 buyer-style "AI Infrastructure" queries
Databricks appears in 529 of 660 prompts (80% shortlist rate) and earns the highest evaluation score at 84 across 99 comparison queries. Weights & Biases (72% shortlist rate) and Google Cloud (70%) round out the top three, while MLflow—an open-source experiment tracker—holds 67% shortlist share. Vector databases Pinecone and Weaviate each clear 57% shortlist penetration but score only 25 and 26 respectively, signaling high visibility with lukewarm AI recommendations.
Hover or click a logo to see brand details
Get weekly AI visibility changes for AI Infrastructure sent to your inbox.
Score shifts, new entrants, citation gaps — every Monday.
Signal by intent
By topic
Top 5 most-cited brands per intent cluster. Brands with zero citations in a topic are not shown.
Evaluation stage
Head-to-head
How often AI cites each brand across uniform category evaluation prompts · median 5/100
Databricks dominates head-to-head queries with an 84 average across 99 comparisons, 14 points above second-place Snowflake (70 across 83 queries). MLflow (69) and Weights & Biases (67) cluster tightly behind Snowflake, while Nvidia sits at 66 despite appearing in 353 prompts. The median evaluation score across all 90 brands is 5, underscoring extreme bifurcation between platform leaders and long-tail niche tools.
Hover or click a logo to see brand details
Each brand's score is the share of category evaluation prompts where AI cited them across all four engines — the same prompt pool for every brand. Brands above the median citation rate have stronger presence in evaluation-stage queries.
Trends
Over the last 12 months
Seven-month score history shows violent swings: April 2026 peaked at 70, May dropped to 34, July rebounded to 75, September settled at 35. Datadog climbed 15 points month-over-month while H2O shed 63, the steepest decline in the category. Small sample sizes (7–35 brands per month) make trend interpretation unstable.
Brands to know
In this category
Databricks
Unified analytics platform owns 80% shortlist rate and 84 evaluation score, the highest in both dimensions. Lakehouse architecture and notebook-based workflows make it the default recommendation for teams consolidating data engineering and ML.
Read brand profile →Weights & Biases
Experiment-tracking specialist reaches 72% shortlist share and 67 evaluation score across 79 queries. Strong model-diversity footprint (4 LLMs cite it) signals broad consensus for ML observability and hyperparameter sweeps.
Read brand profile →MLflow
Open-source MLOps standard achieves 67% shortlist rate and 69 evaluation score despite carrying no vendor lock-in. 892 total mentions across 660 prompts indicate repeated co-recommendation with cloud platforms.
Read brand profile →Snowflake
Data warehouse scores 78 overall but only 70 in head-to-head evaluations, trailing Databricks by 14 points. 62% shortlist rate reflects its strength in structured analytics before teams layer on ML workflows.
Read brand profile →Drill down
Subcategories
Want to know if AI cites your brand for AI Infrastructure?
Free audit. ChatGPT, Perplexity, Gemini, Claude.
Run an audit →