Skip to content

Commit e9d69cb

Browse files
Paul-CornellPaul-Cornell
authored
IBM embedding models; IBM watsonx Orchestrate tool example (#744)
1 parent 64a7d4e commit e9d69cb

File tree

3 files changed

+184
-0
lines changed

3 files changed

+184
-0
lines changed

‎api-reference/workflow/workflows.mdx‎

Lines changed: 8 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -2076,6 +2076,14 @@ Allowed values for `subtype` and `model_name` include the following:
20762076
- `"model_name": "cohere.embed-english-v3"`
20772077
- `"model_name": "cohere.embed-multilingual-v3"`
20782078

2079+
- `"subtype": "ibm"`
2080+
2081+
- `"model_name": "sentence-transformers/all-minilm-l6-v2"`
2082+
- `"model_name": "ibm/granite-embedding-278m-multilingual"`
2083+
- `"model_name": "intfloat/multilingual-e5-large"`
2084+
- `"model_name": "ibm/slate-125m-english-rtrvr-v2"`
2085+
- `"model_name": "ibm/slate-125m-multilingual-rtrvr-v2"`
2086+
20792087
- `"subtype": "togetherai"`
20802088

20812089
- `"model_name": "togethercomputer/m2-bert-80M-32k-retrieval"`

‎docs.json‎

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -304,6 +304,7 @@
304304
"examplecode/tools/vectorshift",
305305
"examplecode/tools/mcp",
306306
"examplecode/tools/mcp-partition",
307+
"examplecode/tools/ibm-orchestrate",
307308
"examplecode/tools/snowflake-streamlit",
308309
"examplecode/tools/crewai",
309310
"examplecode/tools/neo4j-chatbot"

‎examplecode/tools/ibm-orchestrate.mdx‎

Lines changed: 175 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,175 @@
1+
---
2+
title: IBM watsonx Orchestrate
3+
---
4+
5+
[IBM watsonx Orchestrate](https://www.ibm.com/products/watsonx-orchestrate) helps you build, deploy and manage powerful
6+
AI assistants and agents that automate workflows and processes with generative AI.
7+
8+
This article provides a hands-on, step-by-step walkthrough that uses IBM watsonx Orchestrate to
9+
build a simple AI chat app. This chat app relies on data that is stored in an Astra DB or Milvus vector database. This data
10+
is generated by Unstructured and is based on your organization's source documents and semi-structured data.
11+
12+
## Requirements
13+
14+
To use this example, you will need the following:
15+
16+
If you want to connect Unstructured to Astra DB for providing the source data to IBM watsonx Orchestrate, you will need the following:
17+
18+
- An [IBM Cloud account](https://cloud.ibm.com/registration) or [DataStax account](https://astra.datastax.com/signup) account.
19+
- An [Astra DB database](https://accounts.datastax.com/session-service/v1/login).
20+
- To complete this example, you will need the following settings for the Astra DB database:
21+
22+
- The database's API endpoint.
23+
- An application token for the database.
24+
- The name of the target keyspace in the database.
25+
- The name of the target collection in the keyspace.
26+
27+
To get these settings, see the [Astra DB destination connector](/ui/destinations/astradb) documentation.
28+
29+
- Within your Unstructured account, a workflow that contains an Astra DB destination connector.
30+
31+
- Create an [Unstructured account](https://unstructured.io/?modal=try-for-free).
32+
- Create an [Astra DB destination connector](/ui/destinations/astradb) in your Unstructured account.
33+
- Create a [custom workflow](/ui/workflows#create-a-custom-workflow) that contains the Astra DB destination connector in your Unstructured account.
34+
35+
The workflow must generate [embeddings](/ui/embedding). The workflow's **Embedder** node's selected embedding model provider must be **IBM**. The
36+
node's selected embedding model must match the embedding model that you will specify later in **Step 2**.
37+
38+
Be sure to [run the workflow](/ui/workflows#edit%2C-delete%2C-or-run-a-workflow) to have Unstructured generate the data and store it in your Astra DB database.
39+
40+
After you meet the preceding requirements, skip ahead to [Step 1](#step-1%3A-add-an-ibm-watsonx-orchestrate-subscription-to-your-ibm-cloud-account).
41+
42+
If, however, you want to connect Unstructured to Milvus on IBM watsonx.data for providing the source data to IBM watsonx Orchestrate, you will need the following:
43+
44+
- Within your IBM Cloud account, an IBM watsonx.data subscription that contains a Milvus service instance.
45+
46+
- Create an [IBM Cloud account](https://cloud.ibm.com/registration).
47+
48+
To complete this example, you must [create an API key](https://www.ibm.com/docs/en/masv-and-l/cd?topic=cli-creating-your-cloud-api-key) for your IBM Cloud account.
49+
50+
- Create an [IBM watsonx.data subscription](https://cloud.ibm.com/watsonxdata) in your IBM Cloud account.
51+
- Create a [Milvus service instance](/ui/destinations/milvus) within your IBM watsonx.data subscription plan. This instance
52+
must contain a database, a collection, and an index to store and manage the data that is generated by Unstructured.
53+
54+
To complete this example, you will need the following settings for the Milvus service instance:
55+
56+
- The instance's GRPC host value.
57+
- The instance's GRPC port value.
58+
- The name of the target database on the instance.
59+
- The name of the target collection in the database.
60+
- The name of the target index in the collection.
61+
62+
To get these settings, see the [Milvus destination connector](/ui/destinations/milvus) documentation.
63+
64+
- Within your Unstructured account, a workflow that contains a Milvus destination connector.
65+
66+
- Create an [Unstructured account](https://unstructured.io/signup).
67+
- Create a [Milvus destination connector](/ui/destinations/milvus) in your Unstructured account.
68+
- Create a [custom workflow](/ui/workflows#create-a-custom-workflow) that contains the Milvus destination connector in your Unstructured account.
69+
70+
The workflow must generate [embeddings](/ui/embedding). The workflow's **Embedder** node's selected embedding model provider must be **IBM**. The
71+
node's selected embedding model must match the embedding model that you will specify later in **Step 2**.
72+
73+
Be sure to [run the workflow](/ui/workflows#edit%2C-delete%2C-or-run-a-workflow) to have Unstructured generate the data and store it in your Milvus instance's database.
74+
75+
## Step 1: Add an IBM watsonx Orchestrate subscription to your IBM Cloud account
76+
77+
If you already have an IBM watsonx Orchestrate subscription, then skip ahead to [Step 2](#step-2%3A-create-the-chat-app-in-ibm-watsonx-orchestrate)
78+
79+
1. [Log in to your IBM Cloud account](https://cloud.ibm.com/login).
80+
2. On the sidebar, click the **Resource list** icon. If the sidebar is not visible, click the **Navigation Menu** icon to the far left of the
81+
top navigation bar.
82+
3. Click **Create resource**.
83+
4. With **IBM Cloud catalog** selected, search for and select **watsonx Orchestrate**.
84+
5. Complete the on-screen instructions to finish creating the IBM watsonx Orchestrate subscription.
85+
86+
## Step 2: Create the chat app in IBM watsonx Orchestrate
87+
88+
In this step, you use IBM watsonx Orchestrate to create a chat app. This chat app allows you to ask questions about
89+
your organization's documents and semi-structured data. This data is stored in your Astra DB or Milvus vector database and was
90+
generated by Unstructured in a format that is well-suited for your chat app.
91+
92+
1. Open your IBM watsonx Orchestrate subscription, if it is not already open. To do this:
93+
94+
a. [Log in to your IBM Cloud account](https://cloud.ibm.com/login).<br/>
95+
b. On the sidebar, click the **Resource list** icon. If the sidebar is not visible, click the **Navigation Menu** icon to the far left of the
96+
top navigation bar.<br/>
97+
c. In the list of resources, expand **AI / Machine Learning**, and then click the target watsonx Orchestrate subscription.<br/>
98+
d. Click **Launch watsonx Orchestrate**.<br/>
99+
100+
2. If the **Chat** page is not already open, click the **Open the main menu** icon to the far left of the
101+
top navigation bar, and then click **Chat**.
102+
3. Toward the bottom of the sidebar, click **Create new agent**.
103+
4. On the **Create an agent** page, with **Create from scratch** already selected, enter some **Name** and **Description** for the agent, and then click **Create**.
104+
5. Under **Knowledge source**, in the **Start by adding knowledge** tile, click **Choose knowledge**.
105+
106+
If you are using Astra DB for providing the source data to IBM watsonx Orchestrate, then do the following:
107+
108+
1. On the **Select source** page, click **Astra DB**, and then click **Next**.
109+
2. On the **Connect Astra DB** page, specify the following settings:
110+
111+
a. For **URL**, enter the API endpoint for the Astra DB database.<br/>
112+
b. Leave **Astra DB port** blank.<br/>
113+
c. For **API key**, enter the application token for your Astra DB database.<br/>
114+
d. Click **Next**.<br/>
115+
116+
3. On the **Settings details** page, specify the following settings:
117+
118+
a. For **Keyspace**, select the name of the target keyspace in the Astra DB database.<br/>
119+
b. For **Data type**, select **Collection**.<br/>
120+
c. For **Collection**, select the name of the target collection in the database.<br/>
121+
d. For **Embedding mode**, select **Client**.<br/>
122+
e. For **Embedding model**, select the name of the embedding model that matches the one that you specified earlier in your Unstructured workflow.<br/>
123+
f. For **Search mode**, select **Vector**.<br/>
124+
g. For **Title**, enter `record_id`.<br/>
125+
h. For **Body**, enter `content`.<br/>
126+
i. Leave **URL** and **Advanced settings** blank.<br/>
127+
j. Click **Next**.<br/>
128+
129+
4. On the **Description** page, enter some description for the agent, and then click **Save**.
130+
5. In the list of pages, click **Behavior**.
131+
6. At the bottom of the **Behavior** section, turn on **Chat with documents**.
132+
7. In the upper-right corner of the page, click **Deploy**.
133+
8. On the **Pre-deployment summary** page, click **Deploy**.
134+
135+
Skip ahead to [Step 3](#step-3%3A-run-the-chat-app).
136+
137+
If, however, you are using Milvus on IBM watsonx.data for providing the source data to IBM watsonx Orchestrate, then do the following:
138+
139+
1. On the **Select source** page, click **Milvus**, and then click **Next**.
140+
2. On the **Connect Milvus** page, specify the following settings:
141+
142+
a. For **GRPC host**, enter the GRPC host value for the Milvus service instance within your IBM watsonx.data subscription.<br/>
143+
b. For **GRPC port**, enter the GRPC port value for the instance.<br/>
144+
c. For **Choose an authentication type**, select **Basic authentication**.<br/>
145+
d. For **Username**, enter `ibmlhapikey`.<br/>
146+
e. For **Password**, enter the API key for your IBM Cloud account.<br/>
147+
f. Click **Next**.<br/>
148+
149+
3. On the **Select index** page, specify the following settings:
150+
151+
a. For **Database**, select the name of the target database on the Milvus service instance within your IBM watsonx.data subscription.<br/>
152+
b. For **Use Collection or Alias**, select **Collection**.<br/>
153+
c. For **Collection**, select the name of the target collection in the database.<br/>
154+
d. For **Index**, select the name of the target index in the collection.<br/>
155+
e. For **Embedding model**, select the name of the embedding model that matches the one that you specified earlier in your Unstructured workflow.<br/>
156+
f. For **Title**, select **element_id**.<br/>
157+
g. For **Body**, select **text**.<br/>
158+
h. Click **Next**.<br/>
159+
160+
4. On the **Description** page, enter some description for the agent, and then click **Save**.
161+
5. At the bottom of the **Behavior** section, turn on **Chat with documents**.
162+
6. In the upper-right corner of the page, click **Deploy**.
163+
7. On the **Pre-deployment summary** page, click **Deploy**.
164+
165+
## Step 3: Run the chat app
166+
167+
In this step, you ask questions about your organization's source documents and semi-structured data. The chat app then
168+
attempts to answer your questions by searching the related data that Unstructured generated and stored in your Astra DB or Milvus vector database.
169+
170+
1. If the **Chat** page is not already open in IBM watsonx Orchestrate, click the **Open the main menu** icon to the far left of the
171+
top navigation bar, and then click **Chat**.
172+
2. In the sidebar, in the **Agents** list, select the name of the agent that you created in the previous step.
173+
3. In the **Type something** box, enter a question, and then press `Enter`.
174+
4. The agent will provide an answer.
175+
5. Keep asking as many questions as you want to.

0 commit comments

Comments
 (0)