Schift is a company-document-based AI answer infrastructure. It is built for situations where questions about internal materials, customer-facing help content, policies, pricing, or other organizational documents need an answer with supporting evidence rather than a pointer to a file or a person. The product presents that workflow as connecting existing tools and then returning an answer within minutes, alongside the source material that supports it.
Its workflow includes document upload, OCR and layout parsing, chunking and embedding, search and reranking, and answer generation with citations. Supported uploads include PDF, HWP, and image files, and the page says those can be added without format conversion or preprocessing. The OCR and layout parsing step identifies tables, stamps, and handwritten text. Documents are indexed as 1024-dimensional vectors, then searched by a Rust engine that finds candidates in 0.3 milliseconds and a reranker orders the results. Answers cite source sentences and can show which document and page they come from. The site also lists source display, permission-based search, a workflow builder, an AI agent area, OCR, PII, RAG Bench, and Engine Speed.
Example uses named on the page include customer support chatbots, document summarization, report generation, legal and policy Q&A, sales proposal drafting, operational automation, and content drafting in an organization’s own tone. It also describes use for product questions, security review, and pricing-related questions, and says the same engine can serve different departments such as customer service, reports, regulations and law, sales, operations, and content.
Schift is delivered through its website and also offers a document API reference, Python SDK, TypeScript SDK, and a quick-start section. The site is available in English, Korean, and Japanese. Pricing is presented as a starting plan at $100 per month for a repeated-question bundle and a team plan at $300 per month for multiple products and multiple responders, with approval, record-keeping, and permission management. A custom-environment option is listed with security consultation, and the page mentions SSO/SCIM custom integration and security review. The company footer names room821, and the site also links to docs, blog, benchmark, GitHub, company, security, and privacy pages.
langchain-schift sits in PulseGate's Infrastructure & Backend category. It focuses on integrating Schift platform capabilities into LangChain agent workflows for developers. langchain-schift is an open-source project aimed at python developers building AI agent workflows. langchain-schift is open source under the MIT license. It runs on the web and the command line, and it can be self-hosted.
Behind langchain-schift is schift-io, and it first shipped in 2026. Key capabilities include langChain integration, schift connector, and python package.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Same category — not a similarity match