Choose theme
AI Resources
DocETL
DocETL is a Python framework for applying model-powered processing steps to a collection of records. You can extract fields, classify entries and group the results for a summary or further analysis.
The project comes from UC Berkeley research groups. Its companion interface, DocWrangler, lets you work on a pipeline and inspect the output from individual steps. Use this as a first read, not a recommendation. Open the original project before trusting details like terms, limits, privacy, cost, setup, or safety.
What it is
Repeatable steps for a collection of text
Describe the operation in a prompt and specify the fields you want back. Define the pipeline in Python or YAML so the same steps can run over the next batch.
Why it stands out
Work on each record, then on each group
A map step applies an operation to each record. A reduce step combines records that share a key. For support tickets, that means deciding the category for each ticket first and then summarizing what people reported in each category.
Availability
Python package and browser workspace
DocETL is published on PyPI for Python 3.10 or later. DocWrangler has hosted and self-hosted setup paths. The main repository lists the MIT license and provides the source and examples.
Why it matters
What makes it useful
A folder of support tickets may contain the same complaint phrased dozens of ways. The project example assigns a category and priority to each ticket, then writes a summary for each category. Those two steps give you ticket-level labels to inspect and a shorter account of what people reported.
What to know
Where it fits
The ticket example starts with a JSON collection containing ticket text. For a document workflow, this is the stage where you decide what fields or groups you want from the material you have collected. DocETL supports Python and YAML for the pipeline itself, and DocWrangler provides a browser workspace for trying operations and refining prompts against sample output.
Notable points
What stands out
The Python example previews the grouped results using five documents. To check individual labels, inspect sample output from the map step in DocWrangler. If similar tickets land in different categories, you can refine the category instructions before those labels determine which tickets get summarized together.
Before using
What to review
The Python package requires Python 3.10 or later. Hosted model providers require their own credentials. DocETL also documents a local Ollama setup.
The hosted DocWrangler documentation says the service caches results and stores data on its servers. Its setup guide describes a local interface as another option. Check where model requests go separately when choosing the provider for that setup.
Inspect the labels and extracted fields before relying on a group summary. A record filed under the wrong category can change which details the summary reflects.
Reader fit
Who may find it relevant
DocETL is geared toward people who need to repeat the same analysis across many records, such as a team sorting feedback each month. It helps when you can describe the fields or categories you want and can spare time to check examples from the result.
Editorial note
Why LifeHubber lists it
LifeHubber lists DocETL because DocWrangler can export results from any operation to CSV. If a category summary looks wrong after a run, you can export the ticket labels from the preceding step and check which tickets were grouped together. That gives the person reviewing a report a way to trace whether a missing complaint was filed under a different category or dropped from the summary.
Source links
Source materials
Reader note
Before relying on this entry
LifeHubber lists entries to help readers inspect AI projects, not to endorse them or prove they are safe, suitable, accurate, maintained, or right for a specific use. We do not verify every entry in depth. Before relying on anything listed, review the original materials, terms, privacy practices, limits, and risks that matter for your situation.
More in Ecosystem
Keep browsing this category
Explore more AI ecosystem resources.
Docling
docling-project/docling
A document conversion toolkit with structured text, tables and layout, native chunking for retrieval, Markdown and JSON exports, and local model configuration.
Xberg
xberg-io/xberg
A Rust-based engine for extracting text, tables and metadata from mixed files, with language bindings, CLI and server interfaces, configurable OCR and retrieval preparation.
SwarmLLM
Nehanth/swarmllm
A browser-based distributed inference project that divides model layers across devices, runs them with WebGPU, and passes intermediate activations over direct WebRTC connections, with hosted and local static-server paths.
For project maintainers
Listed here? You can use the badge.
If you maintain a project with a current LifeHubber listing, you may add the optional “Listed on LifeHubber AI Resources” badge to its README, docs, or website. No introduction or permission request is needed.
