Sim Search

Product Introduction

1. Definition: Sim Search is a knowledge graph creation and search engine feature within Sim, an open-source AI workspace platform. Technically, it is a semantic search and data structuring tool that transforms unstructured personal and company data into a connected, queryable knowledge graph.
2. Core Value Proposition: It exists to solve the problem of fragmented, unstructured data in enterprises, enabling AI agents and human users to access and reason over all organizational knowledge from a single, intelligent source. Its primary value is turning disparate data sources into a unified, AI-ready knowledge base for enhanced agent performance and decision-making.

Main Features

1. Knowledge Graph Creation & Auto-Sync: Sim Search automatically ingests data from connected sources like Notion, Google Drive, Confluence, Slack, GitHub, and SharePoint. It doesn't just store documents; it processes and structures the information, identifying entities and relationships to build a dynamic knowledge graph. This is powered by embedding models and graph database technologies, creating a semantic layer over raw data.
2. Semantic & Vector Search: Unlike traditional keyword search, Sim Search uses vector embeddings to understand the contextual meaning of queries. This allows users and AI agents to perform nuanced searches like "find projects related to Q4 budget risks" and receive results based on conceptual relevance, not just matching words. This leverages large language models (LLMs) for query understanding and vector similarity search.
3. Structured Data Integration with Tables: Beyond document search, Sim Search integrates with Sim's native Tables feature. This allows the knowledge graph to incorporate and reference structured data (e.g., CRM leads, project statuses). Agents can query across both unstructured knowledge and structured datasets in a single operation, enabling complex workflows like enriching a lead with relevant company documentation automatically.
4. Centralized Governance & Audit for Knowledge: As part of the Sim platform, access to Sim Search is governed by centralized permission groups and SSO. All searches and data accesses are logged within Sim's audit records and run tracing, providing full visibility into how organizational knowledge is being used by both humans and automated agents.

Problems Solved

1. Pain Point: Organizational knowledge silos and data fragmentation. Critical information is locked in separate tools (Slack, Drive, tickets, databases), making it inaccessible for AI agents and inefficient for teams to find, leading to duplicated work and inconsistent decisions.
2. Target Audience: AI/ML Engineers & Developers building context-aware agents; Operations & Support Managers needing to automate customer service with accurate internal knowledge; Enterprise Architects & IT Leaders tasked with governing AI and data access; Sales & Marketing Teams requiring quick access to product and client data.
3. Use Cases: A support agent workflow that automatically searches the knowledge graph (product docs, past tickets) to draft accurate replies. A sales enrichment agent that pulls company background from synced Confluence pages when a new lead arrives. An R&D team using semantic search across all internal research notes and GitHub repositories to accelerate development.

Unique Advantages

1. Differentiation: Unlike standalone vector databases (e.g., Pinecone) or search appliances, Sim Search is natively integrated into a full AI agent orchestration platform. This eliminates the need to build and maintain a separate data pipeline for AI. Compared to generic enterprise search, it is built specifically for agentic consumption with built-in governance and workflow triggers.
2. Key Innovation: The seamless, bi-directional integration between the Knowledge Graph (Sim Search), Agent Workflow Engine, and Structured Data (Tables). This creates a closed-loop system where agents can both query knowledge and write structured results back, updating the organization's "memory." The open-source, self-hostable nature also provides unique data sovereignty and customization advantages for enterprises.

Frequently Asked Questions (FAQ)

1. How does Sim Search handle data privacy and security? Sim Search inherits the enterprise-grade security of the Sim platform, including SOC 2 Type II and ISO 27001 compliance. Data can be self-hosted in your own cloud, and access is controlled via SSO, SCIM, and granular permission groups. All data ingestion and queries are audited.
2. What data sources can Sim Search connect to for building the knowledge graph? It offers hundreds of pre-built connectors for popular SaaS tools including Notion, Google Drive, Confluence, Slack, GitHub, SharePoint, and many databases. It supports auto-syncing to keep the knowledge graph continuously updated.
3. Can Sim Search be used via an API for custom applications? Yes, Sim Search functionality is accessible via the Sim API and CLI, allowing developers to programmatically query the knowledge graph from external applications, custom dashboards, or other AI systems.
4. How is Sim Search different from using ChatGPT with a file upload? While ChatGPT can analyze individual files, Sim Search creates a persistent, searchable knowledge graph from all company data sources. It provides centralized governance, maintains data freshness through syncs, and allows complex, multi-step agent workflows to use this knowledge automatically, at scale.
5. Is coding required to set up and use Sim Search? No, the initial setup—connecting data sources, configuring syncs, and performing searches—can be done through Sim's visual interface and chat. However, the platform also offers a full CLI and API for developers who require advanced customization and integration.