notifications
News
For Power Users
PromptPerfect
Premier tool for prompt engineering
SceneXplain
Leading AI solution for image captions and video summaries
BestBanner
Blog to banner, without the prompts!
JinaChat
More modality, longer memory, less cost
Rationale
Ultimate AI decision-making tools


For Developers
DocArray
The data structure for multimodal data
Jina
Build multimodal AI applications on the cloud
Finetuner
Fine-tune embeddings on domain specific data for better search quality
CLIP-as-service
Embed images and sentences into fixed-length vectors with CLIP
JCloud
Deploy a local project as a cloud service. Radically easy, no nasty surprises.
LangChain-Serve
Langchain apps on production with Jina & FastAPI
VectorDB
A Python vector database you just need - no more, no less
Executor Hub
Share and discover building blocks for multimodal AI applications
DALL-E Flow
A human-in-the-Loop workflow for creating HD images from text
DiscoArt
Create compelling Disco Diffusion artworks in one line of code
ThinkGPT
Agent techniques to augment your LLM and push it beyond its limits
DevGPT
Your virtual development team
RunGPT
An open-source cloud-native of large multimodal models serving framework
Jerboa
An experimental finetuner for open-source LLMs


For Enterprises
Finetuner+
Empower your enterprise with on-premise finetuning solutions
Inference
State-of-the-art multimodal models available for inference
Jina AI Cloud
Cloud hosting platform for multimodal AI applications
Searchscape
Navigate, interact, refine: reimagine product discovery
Semantica
Bridging the semantic gap in your existing search infrastructure


Company
Contact sales
About us
Join us
open_in_new
Open day
Intern program


arrow_circle_leftBack to Newsroom
Tech blog
Gold Retriever: Let ChatGPT talk to your data
Saba Sturua, Alex C-G
May 23, 2023 • 3 minutes read

In the realm of AI, language models like ChatGPT have revolutionized how we interact with machines. However, as efficient as these models are, they encounter certain limitations. One key hurdle is their inability to seamlessly integrate and reason with new, personalized, or up-to-date data, which is often critical for users.

Out-of-date data

Language models have a knowledge cut-off, beyond which they aren't informed about new world events, scientific discoveries, or technological breakthroughs. They may also lack depth in specific knowledge domains, struggle with real-time data, and produce outputs that might need fact-checking for accuracy. Additionally, users desire AI interactions tailored to their profiles and control over their data, which requires the AI to access user-specific data.

Enter "Gold Retriever", a powerful open-source tool powered by Jina and DocArray, that addresses these very needs.

GitHub - jina-ai/GoldRetriever: Create and host retrieval plugins for ChatGPT in one click
Create and host retrieval plugins for ChatGPT in one click - GitHub - jina-ai/GoldRetriever: Create and host retrieval plugins for ChatGPT in one click
GitHubjina-ai

It equips users to create custom plugins that enable ChatGPT to interact with their data, overcoming inherent limitations. Gold Retriever acts as a bridge linking ChatGPT with the vast and dynamic world of data, making AI interactions more meaningful, accurate, and personal.

Gold Retriever in Action: Study Assistant

To better illustrate how Gold Retriever works, let's consider a practical scenario. Imagine you're a student seeking assistance from ChatGPT with your coursework. However, your course materials exceed ChatGPT's prompt limit, preventing you from getting the help you need. That's exactly where Gold Retriever comes in handy.

In this example, we'll use resources from MIT's open course - Ecology I: The Earth System.

Ecology I: The Earth System | Civil and Environmental Engineering | MIT OpenCourseWare
We will cover fundamentals of ecology, considering Earth as an integrated dynamic system. Topics include coevolution of the biosphere, geosphere, atmosphere and oceans; photosynthesis and respiration; the hydrologic, carbon and nitrogen cycles. We will examine the flow of energy and materials throug…
MIT OpenCourseWare

The first step is to gather all relevant text files in one directory, which will serve as your data source. Next, you'll need to create a Gold Retriever plugin. Here's how:

Install Gold Retriever via pip with the command:

pip install goldretriever

Now, deploy a plugin providing your OpenAI key:

goldretriever deploy --key <your openai key>

After deployment, make a note of the "Gateway (HTTP)" URL and the Bearer token, as you'll need them to incorporate the plugin into ChatGPT.

╭──────────────────────── 🎉 Flow is available! ────────────────────────╮
│                                                                       │
│   ID               retrieval-plugin-<plugin id>                       │
│   Gateway (Http)   https://retrieval-plugin-<plugin id>.wolf.jina.ai  │
│   Dashboard        https://dashboard.wolf.jina.ai/flow/<plugin id>    │
│                                                                       │
╰───────────────────────────────────────────────────────────────────────╯
Bearer token: <your bearer token>

Next, let's index the data we've gathered:

goldretriever index --data <dir_name>

With the data indexed successfully, the plugin can now be integrated into ChatGPT. Navigate to plugins, choose "Develop your own plugin", and provide the "Gateway (HTTP)" URL and Bearer token from the deployment phase. Voila! Your plugin is now installed and ready for a test run!

With Gold Retriever at your disposal, there are numerous ways to use ChatGPT to assist with coursework. A fun and effective studying strategy is to create flashcards from course materials, and you can enlist ChatGPT's help for this task:

0:00
/

ChatGPT looks into your notes, extracts relevant material about the chosen topic (Thermodynamics, in this instance), and creates comprehensive flashcards based on the information. It can also provide further clarification on certain concepts:

0:00
/

ChatGPT first identifies the example you're seeking and then responds with a detailed explanation.

This case study showcases just one of the countless applications for Gold Retriever. If you have textual data that you'd like ChatGPT to remember and use, Gold Retriever is your go-to tool.

Be sure to watch the video for the full tutorial:

For more information, head over to the Gold Retriever repository and join the discussion on Discord. Happy coding!


Categories:
Tech blog

Learn more
SceneXplain's Image-to-JSON: Extract Structured Data from Images with Precision
Pushing the boundaries of visual AI, we're thrilled to unveil SceneXplain's Image-to-JSON feature. Dive into a world where images aren't just seen, but deeply understood, translating visuals into structured data with unparalleled precision.
Engineering Group
September 14, 2023 • 6 minutes read
Embeddings: The Swiss Army Knife of AI
Embeddings are an essential technology for modern AI. This article explains what they are and how they work.
Scott Martens
September 13, 2023 • 10 minutes read
PromptPerfect's LLM-as-a-Database is Blurring the Boundaries of Search vs. Generation
Diving into PromptPerfect's LLM-as-a-database: a middle ground between few-shot prompts and RAG. Its nuanced approach outshines GPT-4 in complex queries. Evolution in action.
Engineering Group
September 11, 2023 • 7 minutes read
Offices
location_on
Berlin, Germany (HQ)
Ohlauer Str. 43 (1st floor), zone A, 10999 Berlin, Germany
Geschäftsanschrift: Leipziger str. 96, 10117 Berlin, Germany
location_on
Beijing, China
Level 5, Building 6, No.48 Haidian West St. Beijing Haidian, China
location_on
Shenzhen, China
402, Floor 4, Fu'an Technology Building, Shenzhen Nanshan, China
© Jina AI GmbH 2020-2023. All rights reserved.
email
Privacy PolicyTerms and ConditionsPrivacy Settings