Skip to content

Search Infrastructure Arena

Search Infrastructure arenaBuyer checklist

Every requirement we judge search infrastructure products against, as a ready-to-send RFP checklist — with each item's priority, why it matters, and how the top-ranked products score on it today.

53 requirements · 13 themes · verdicts for 5 products · updated 2026-09-16 · priorities mirror the story weights our scoring uses (methodology)

Procurement report →
Show the markdown export
# Search Infrastructure — buyer checklist (RFP)

Derived from ProductArena's evidence-graded user-story taxonomy for Search Infrastructure: 53 judged requirements. Priorities mirror story weights (3 = must-have, 2 = should-have, 1 = nice-to-have).

## Agent search

- [ ] **[must-have]** My coding agent can create an index, add documents, and run queries end to end — through the API, CLI, or MCP without touching a dashboard
- [ ] **[should-have]** Agents can use my search indexes as a tool — an MCP server or tool-calling surface exposes query, analytics, and index operations

## Agenticness

- [ ] **[must-have]** Plug MCP servers into this product so it can use their tools
- [ ] **[must-have]** Connect an agent via an official MCP server
- [ ] **[must-have]** Drive the product through a documented public API
- [ ] **[must-have]** Delegate tasks to a built-in AI assistant inside the product
- [ ] **[should-have]** Point an agent at llms.txt or agent-oriented docs
- [ ] **[should-have]** Run the product headlessly / in CI for automation
- [ ] **[should-have]** Use an official CLI
- [ ] **[should-have]** Issue scoped/least-privilege API credentials for an agent
- [ ] **[should-have]** Build against official SDKs
- [ ] **[should-have]** Subscribe to events via webhooks
- [ ] **[should-have]** Get AI-generated insights and suggestions from my data inside the product
- [ ] **[should-have]** Set up automations that run autonomously in the background
- [ ] **[should-have]** Operate the product with natural-language commands
- [ ] **[should-have]** Explore an interactive API reference with runnable examples
- [ ] **[should-have]** Download a machine-readable API spec (OpenAPI or equivalent)
- [ ] **[should-have]** Rely on versioned APIs with a documented deprecation policy
- [ ] **[nice-to-have]** Test against a sandbox environment without touching production data

## Ai search

- [ ] **[must-have]** Run hybrid search — semantic vector similarity fused with keyword matching — in a single query
- [ ] **[should-have]** Use built-in or managed embedders so documents and queries are vectorized without running my own embedding pipeline
- [ ] **[nice-to-have]** Power RAG and conversational answers on top of my indexes with documented retrieval or answer APIs

## Automation depth

- [ ] **[must-have]** Define rules that trigger actions automatically on events
- [ ] **[should-have]** Perform bulk operations across many items at once
- [ ] **[should-have]** Schedule recurring jobs or workflows
- [ ] **[nice-to-have]** Version, review, and roll back my automations

## Developer experience

- [ ] **[must-have]** Create an index, add documents, and run my first search within minutes of starting the quickstart
- [ ] **[should-have]** Official SDKs cover my language and framework, kept current across JavaScript, Python, PHP, Ruby, Go, and more
- [ ] **[should-have]** Official UI component libraries let me assemble a full search interface — box, results, facets, pagination — without building it from scratch

## Indexing pipelines

- [ ] **[should-have]** Bulk-import millions of documents quickly, with async task tracking to know when indexing completes
- [ ] **[should-have]** Document adds, updates, and deletes become searchable in near real time without a full reindex
- [ ] **[nice-to-have]** Ingest content with an official crawler or connectors instead of writing my own indexing pipeline

## Openness

- [ ] **[must-have]** Export all of my data in open formats and leave
- [ ] **[must-have]** Self-host the core product
- [ ] **[should-have]** Do everything through the API that I can do in the UI
- [ ] **[should-have]** Read the product's source under an open license

## Operations scale

- [ ] **[must-have]** Self-host the full engine — same features as the hosted product — on my own infrastructure
- [ ] **[should-have]** Built-in analytics show top queries, no-result queries, and click-through so I know what users search for and miss
- [ ] **[should-have]** Documented scaling paths — clustering, replication, high availability — carry me from prototype to hundreds of millions of documents

## Pricing plans

- [ ] **[must-have]** Published per-unit pricing (searches, records, or nodes) lets me predict what search will cost before committing
- [ ] **[should-have]** Costs stay predictable as records and query volume grow — no surprise per-request cliffs

## Privacy posture

- [ ] **[must-have]** Prevent my data from being used to train AI models
- [ ] **[should-have]** Choose where my data is stored (region/residency)
- [ ] **[should-have]** Control data retention and deletion
- [ ] **[should-have]** Opt out of telemetry and usage tracking

## Relevance tuning

- [ ] **[should-have]** Define synonyms and curate results — pin, boost, or hide specific hits for specific queries
- [ ] **[should-have]** Shape relevance with custom ranking rules and business signals (popularity, recency, margin) beyond textual matching
- [ ] **[nice-to-have]** Inspect ranking scores or explanations to understand exactly why a result ranked where it did

## Search experience

- [ ] **[must-have]** Deliver as-you-type instant search with millisecond responses so results update on every keystroke
- [ ] **[must-have]** Searches tolerate typos and misspellings out of the box, with tunable rules for when and how fuzzy matching applies
- [ ] **[should-have]** Serve query suggestions and autocomplete backed by real search traffic or a suggestions index
- [ ] **[should-have]** Build faceted navigation — filters with live counts across categories, ranges, and attributes — from a single query

## Security multitenancy

- [ ] **[should-have]** Scoped or tenant tokens restrict each end user's searches to their own documents without separate indexes per user

---

Source: https://ultrametric.ai/productarena/arena/search-infra (evidence-graded verdicts for 5 products) · methodology: https://ultrametric.ai/productarena/methodology

Chips show the top 5 ranked products' current verdict on each requirement — ✓ full · ~ partial · ! disputed · — none · n/a not applicable.

Agent search — stories about agent search in this arenaAgent search· 2 items

Stories about agent search in this arena

  • ai-native userMy coding agent can create an index, add documents, and run queries end to end — through the API, CLI, or MCP without touching a dashboard

    Core requirement — weighs 3× in arena scoring · 3 of 5 products fully deliver this today

    must-have
  • ai-native userAgents can use my search indexes as a tool — an MCP server or tool-calling surface exposes query, analytics, and index operations

    Important, not disqualifying — weighs 2× in arena scoring · all 5 products fully deliver this today

    should-have

Agenticness — how well agents can access and operate the productAgenticness· 17 items

How well agents can access and operate the product

Ai search — stories about ai search in this arenaAi search· 3 items

Stories about ai search in this arena

  • developerRun hybrid search — semantic vector similarity fused with keyword matching — in a single query

    Core requirement — weighs 3× in arena scoring · all 5 products fully deliver this today

    must-have
  • developerUse built-in or managed embedders so documents and queries are vectorized without running my own embedding pipeline

    Important, not disqualifying — weighs 2× in arena scoring · 2 of 5 products fully deliver this today

    should-have
  • developerPower RAG and conversational answers on top of my indexes with documented retrieval or answer APIs

    Differentiator, not a dealbreaker — weighs 1× in arena scoring · 2 of 5 products fully deliver this today

    nice-to-have

Automation depth — how much of the product can run unattendedAutomation depth· 4 items

How much of the product can run unattended

Developer experience — stories about developer experience in this arenaDeveloper experience· 3 items

Stories about developer experience in this arena

  • developerCreate an index, add documents, and run my first search within minutes of starting the quickstart

    Core requirement — weighs 3× in arena scoring · 2 of 5 products fully deliver this today

    must-have
  • developerOfficial SDKs cover my language and framework, kept current across JavaScript, Python, PHP, Ruby, Go, and more

    Important, not disqualifying — weighs 2× in arena scoring · 1 of 5 products fully deliver this today

    should-have
  • developerOfficial UI component libraries let me assemble a full search interface — box, results, facets, pagination — without building it from scratch

    Important, not disqualifying — weighs 2× in arena scoring · 1 of 5 products fully deliver this today

    should-have

Indexing pipelines — stories about indexing pipelines in this arenaIndexing pipelines· 3 items

Stories about indexing pipelines in this arena

  • platform-engineerBulk-import millions of documents quickly, with async task tracking to know when indexing completes

    Important, not disqualifying — weighs 2× in arena scoring · no product fully delivers this yet

    should-have
  • developerDocument adds, updates, and deletes become searchable in near real time without a full reindex

    Important, not disqualifying — weighs 2× in arena scoring · no product fully delivers this yet

    should-have
  • founderIngest content with an official crawler or connectors instead of writing my own indexing pipeline

    Differentiator, not a dealbreaker — weighs 1× in arena scoring · 1 of 5 products fully deliver this today

    nice-to-have

Openness — open source, data portability, and self-hosting storiesOpenness· 4 items

Open source, data portability, and self-hosting stories

Operations scale — stories about operations scale in this arenaOperations scale· 3 items

Stories about operations scale in this arena

  • platform-engineerSelf-host the full engine — same features as the hosted product — on my own infrastructure

    Core requirement — weighs 3× in arena scoring · 1 of 5 products fully deliver this today

    must-have
  • founderBuilt-in analytics show top queries, no-result queries, and click-through so I know what users search for and miss

    Important, not disqualifying — weighs 2× in arena scoring · 1 of 5 products fully deliver this today

    should-have
  • platform-engineerDocumented scaling paths — clustering, replication, high availability — carry me from prototype to hundreds of millions of documents

    Important, not disqualifying — weighs 2× in arena scoring · no product fully delivers this yet

    should-have

Pricing plans — plan structure and value — what each tier costs and what it unlocksPricing plans· 2 items

Plan structure and value — what each tier costs and what it unlocks

Privacy posture — data-handling and privacy storiesPrivacy posture· 4 items

Data-handling and privacy stories

Relevance tuning — stories about relevance tuning in this arenaRelevance tuning· 3 items

Stories about relevance tuning in this arena

Search experience — stories about search experience in this arenaSearch experience· 4 items

Stories about search experience in this arena

Security multitenancy — stories about security multitenancy in this arenaSecurity multitenancy· 1 item

Stories about security multitenancy in this arena

Full evidence behind every verdict lives on the arena page and each product page — chips above deep-link straight to the judged story.