Organization repository on GitHub
Long-document search with Jev and PageIndex
- Forks
- 7
- Last pushed
- Updated Oct 2026
Repo stats from the GitHub API, cached Oct 2026.Homepage
Jev directory
CommunityJev Integrations & Adapters
Jev Doc Search
Long-document search that pairs Jev's Choice primitive with PageIndex's hierarchical tree: instead of a vector DB or embeddings, Jev picks a node and then a child, so even a 300-page report can be searched a handful of options at a time.
- Stars
- 154
- Language
- Python
- License
- APACHE-2.0
- Reading time
- 2 min
- Published by
- Community
- Author
- VectifyAI
- Last verified
- Oct 2026
Highlights
- Frames page search as one Choice question with one option per page, described by that page's text.
- PageIndex removes the read-at-once limit by turning the document into a hierarchical tree.
- Advertises no vector DB or embeddings; the core is a single typesafe.system_one call with model jev-1.13.0.
- Short-PDF example: python page_search.py handbook.pdf with a vacation-days question prints the likely page.
- Apache-2.0; 153 stars, 7 forks, created 2026-09-30, last pushed 2026-10-03.
Quickstart
bash
python page_search.py handbook.pdf "How many vacation days do new employees get?"Watch out
README reviewed 2026-10-17; the Setup section with install and API-key steps is referenced but not included in the digest excerpt, and the 300-page report is an illustration rather than a measured benchmark.
Questions about this entry
Short answers drawn from the linked project or page.
An example of long-document search with Jev and PageIndex. Instead of a vector database or embeddings, it frames page search as a multiple-choice question and lets Jev's Choice pick the page most likely to answer your question.
Each page becomes one option described by its own text. Jev reads every option and returns a probability for each page, and the most likely page is the answer. The README calls this flat page search: one Choice with one option per page.
PageIndex removes the limit by turning the document into a hierarchical tree representation. Jev picks a node, then one of its children, and so on down the tree, choosing among a handful of options each time however long the document is.
Run python page_search.py handbook.pdf "How many vacation days do new employees get?" and it prints the most likely page and its text, according to the README's short-PDF example. Setup for installing and API keys is described in a section the excerpt does not include.
It builds a Choice with instructions and one criteria entry per page, calls typesafe.system_one with the question as state, and reads r.answers["page"].choice, for example p7, as the most likely page.
Yes. The README points to its Setup section for installing and API keys, and the example uses the typesafe-sdk against model jev-1.13.0, so live searches call TypeSafe Jev rather than running a local model.
The digest excerpt shows no accuracy or latency benchmarks, and the 300-page report is used as an illustration rather than a measured result. It is Apache-2.0 with 153 stars and 7 forks, created 2026-09-30 and last pushed 2026-10-03.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
X: A RAG loop where Jev decides what is real
X: LinkedIn bookmarks search
X: Measuring what Jev does in a RAG pipeline
X: Jev-powered reranking for RAG. Likitd_’s jev-ranker uses Jev to rerank results in a RAG pipeline.
More Jev resources like this
A GitHub App that routes pull-request review work through Jev: it ingests the webhook, normalizes the PR context, gets a LOW, NORMAL or SPECIALIST decision, and a deterministic policy layer maps it to auto-approval, a review request or a check-run update.
Jev Integrations & Adapters#community#catalog-import#logicrw
Community
hexyedev/jevprPython
hexyedev/jevprPythonLiteLLM's proxy adds TypeSafe guardrail hooks and a Jev complexity classifier for policy judgments and request routing across 100+ model providers.
Jev Integrations & Adapters#community#python#gateway
Community
BerriAI/litellmPython
BerriAI/litellmPythonVolcengine's agent context database that unifies memory, knowledge RAG and skills under a viking:// virtual filesystem; it ships a Jev rerank client that scores candidate documents against api.typesafe.ai and treats the probability as relevance.
Jev Integrations & Adapters#community#agent-memory#rag
Community
volcengine/OpenVikingPython
volcengine/OpenVikingPythonAlibaba's multi-agent platform, which implements native TypeSafe Jev classification through JevClassifierModel for binary, choice and score routing across agent pipelines. Python 3.11+, Apache-2.0, with a FastAPI agent service and web UI.
Jev Integrations & Adapters#community#multi-agent#framework
Community
agentscope-ai/agentscopePython
agentscope-ai/agentscopePythonGraph RAG framework over Milvus vector search with no graph database required: Jev Noul judgments select graph relations and rerank full source passages alongside vector-search candidates, reaching 86.07% average Recall@5 on 1,000 MuSiQue and 1,000 2Wiki queries in the two-stage evaluation.
Jev Integrations & Adapters#community#rag#retrieval
Community
zilliztech/vector-graph-ragPython
zilliztech/vector-graph-ragPythonA Python library that runs LLM and Jev calls as Polars expressions: parallel inference across providers, Pydantic-validated outputs as struct columns, and a native TypeSafe System One layer for per-row decision questions.
Jev Integrations & Adapters#community#catalog-import#logicrw
Community
pnthn-ai/polar_llamaPython
pnthn-ai/polar_llamaPythonFrom the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.