1. Google Cloud console, from left menu Agent Platform > Studio
2.left, select + New > Chat
3. Rename from Untitled Prompt to Insurance Risk Summary - Prototype
4. System instructions box, copy paste above : You are an expert AI assistant for an insurance underwriting department. Your primary goal is to help underwriters by accurately and concisely summarizing client information and highlighting potential risk factors. Maintain a professional and objective tone. Focus only on the information provided in the prompt. Do not invent details.
5. In the main prompt area, paste the following: Customer Notes for 'SafeHarbor Warehousing': "The applicant is seeking coverage for their 50,000 sq ft warehouse. The business is 5 years old. The building is a concrete tilt-up structure, originally built in 2010. They store a variety of non-hazardous dry goods. Fire safety measures include a full sprinkler system, a centrally monitored fire alarm, and documented annual inspections by a certified third party. Security measures include a 24/7 centrally monitored burglar alarm, comprehensive security camera coverage of the interior and exterior, a fully fenced perimeter, and nightly patrols by a contracted security guard service. The company reports no major property or liability losses in their 5-year history. They have specifically asked to ensure their new automated shelving and retrieval system, installed last month, is adequately covered under the policy." Your Task: 1. Briefly summarize the key details of the 'SafeHarbor Warehousing' business and its existing safety measures. 2. Based *only* on the notes provided, identify any immediate questions an underwriter should ask or potential risk factors they should consider further. Present the summary first, then the questions/risk factors as bullet points.
6. Model settings: Model > Gemini > model_name
7. Global for Region.
8. Save Submit
9. Deploy icon > Cloud Run > Deploy as app.
10. Acknowledgement and Click Create App.
11. Wait 5-6 minutes saved then close.
12. Deploy > Cloud Run > Open app
13. You should see a new page titled Welcome to Vertex AI Gen AI App! with your prompt title, Insurance Risk Summary - Prototype, displayed.
14. Chatbot section, in the "Type a message...
New Customer Inquiry: "Applicant 'Coastal Goods Delivery' has a fleet of 10 delivery vans, all equipped with GPS and telematics. They operate within a 100-mile radius of their depot. Drivers undergo annual safety training. They had one minor fender bender last year, no injuries, $1500 damages. What are the primary risk considerations?" Please summarize key points and identify potential risks.
15.Yes doneeeeeee be happy Check Progress must be 20%
Task 2. Design effective prompts-> Exit without saving then click Continue.
2A. This is called Zero-shot prompting
1.Task 1, close that browser tab to return to the Google Cloud console.
2.Top left, select New Chat.
3. Top right, click Untitled Prompt and rename/name it: Insurance Claim Data Extraction.
4.System instructions box=:
You are an AI assistant specializing in parsing and extracting specific data points from unstructured insurance claim notifications.
Your goal is to identify and list key information accurately.
If a piece of information is not found, clearly state "Not found".
Output the extracted information in a key: value format, with each key on a new line.
5. Main prompt area, paste =
Claim Notification Received:
"Hi team, just got a call from Mrs. Eleanor Vance, policy #POL458892. She reported a kitchen fire that occurred on May 12th, 2025, around 3 PM. The main damage seems to be to the oven and surrounding cabinets. She mentioned smoke damage in the kitchen and dining area too. She thinks the total damage might be around $7,500. Her contact is 555-0123. No injuries reported, thankfully."
Extract the following:
- Policy Number
- Claimant Name
- Date of Loss
- Time of Loss
- Type of Loss
- Brief Description of Damage
- Estimated Loss Amount
- Injuries Reported
6.Model settings:
Temperature to 0.1
Output token limit to a reasonable number, like 1024.
Ensure the Region is Global.
Submit (Enter)
Task 2B.This is called : Few-Shot prompting
1. On the top left, select New Chat.
2. On the bottom of the Prompt section, click the + button.
3. From the pop-up menu, select Example.
This will open a new window where you can add examples for the prompt.
In the "Examples" interface that appears:
4. For the INPUT field of your first example, paste the following unstructured note:
Claim Notification Received:
"Email from John Sterling (policy POL77521) re: water damage at his shop. Happened sometime last night, May 10th, 2025. A pipe burst in the ceiling. Stockroom is flooded, some damage to inventory. He's not sure on the cost yet, maybe $5k-$10k? No one was there, so no injuries."
Extract the following:
- Policy Number
- Claimant Name
- Date of Loss
- Time of Loss
- Type of Loss
- Brief Description of Damage
- Estimated Loss Amount
- Injuries Reported
5.For the OUTPUT field of your first example, paste the following perfectly formatted extraction:
Policy Number: POL77521
Claimant Name: John Sterling
Date of Loss: May 10th, 2025
Time of Loss: Night
Type of Loss: Water damage
Brief Description of Damage: Pipe burst in ceiling, stockroom flooded, some damage to inventory.
Estimated Loss Amount: $5,000 - $10,000
Injuries Reported: No
6. Click the Add examples button to save this example and return to the main prompt.
Task 2C. 7. Re-add System Instructions: Since clearing the prompt also cleared the system instructions, paste them again into the System instructions box=:
You are an AI assistant specializing in parsing and extracting specific data points from unstructured insurance claim notifications.
Your goal is to identify and list key information accurately.
If a piece of information is not found, clearly state "Not found".
Output the extracted information in a key: value format, with each key on a new line.
8. Provide the New Input and the Prompt:
In the area labeled {Input} Write value here, paste the original claim notification for Eleanor Vance that you want the model to process now=:
Claim Notification Received:
"Hi team, just got a call from Mrs. Eleanor Vance, policy #POL458892. She reported a kitchen fire that occurred on May 12th, 2025, around 3 PM. The main damage seems to be to the oven and surrounding cabinets. She mentioned smoke damage in the kitchen and dining area too. She thinks the total damage might be around $7,500. Her contact is 555-0123. No injuries reported, thankfully."
9.Enter the following instruction in Type a prompt=:
Extract the following data points from the provided claim notification:
- Policy Number
- Claimant Name
- Date of Loss
- Time of Loss
- Type of Loss
- Brief Description of Damage
- Estimated Loss Amount
- Injuries Reported
10. Click on the Model settings:
Ensure the model_name model is selected.
Set Temperature to 0.1 (for more factual, less creative extraction).
Set Output token limit to a reasonable number, like 1024.
Ensure the Region is Global.
11.Click the Submit (Enter)
Task 2 Experinting: Experimenting with prompt configurations:
Now, you'll explore how different parameters in the Model settings can affect the model's response.
Ensure your "Insurance Claim Data Extraction" prompt with the few-shot example is active.
First, create a new prompt to use for these experiments:
1. On the top left, select New Chat.
2. Title the prompt Insurance Story.
3.In the prompt text box, paste the following=:
Write the *first paragraph* of a short story about a homeowner who just used a futuristic AI insurance app to file a claim. The claim was for a bizarre and unexpected incident.
4.Experiment with model settings
5. Temperature:
Lower values (e.g., 0.0-0.2) make the output more focused
Higher values (e.g., 1.5-2)
6. Try it: Change
Temperature to 1.5. Click Submit (Enter) and note any changes.
Then, change Temperature back to 0.1, use the same prompt, and click Submit (Enter) again.
7.Token Limit:
Try it:
SET token limit to a very small number, like 500.
Reset it to the max (default) length 65535.
observe the truncated output.
8. Top-P:value of 1.0 considers all tokens. Lowering Top-P (e.g., to 0.8) makes the output more focused, similar to lowering temperature.
9.Thinking Budget:Auto,Off,Manual
10.Structured output:like JSON
11.Grounding: Google Search: Connects public information.
12.Grounding: Your data:Agent Platform Search or RAG Engine) to answer context-specific questions
Task 3:Engineer and manage prompts
1.On the top left, select New Chat. This will bring you to a new prompt editor page.
2.Name this new prompt : Insurance Risk Factor Identification.
3. Set up this simple base prompt:
4.In the System instructions box, enter=:
You are an insurance risk analyst assistant. Your task is to identify potential risk factors from a given scenario. Be concise.
5. In the main prompt area, paste the following=:
Scenario:
"The applicant, 'The Fiery Grill,' is a new upscale restaurant specializing in wood-fired oven pizzas and open-flame grilling. They have installed a brand new, custom-built fire suppression system for their cooking area, but it has not yet been certified by a third party. The restaurant plans to feature live acoustic music on weekend evenings and has a small, raised stage area. They also want to offer valet parking."
Based on this scenario, list three primary risk factors an underwriter should consider.
6.Click on the Model settings:
Temperature: 0.2
Region: Global
Enter save (Auto save may works)
7.Now, with your Insurance Risk Factor Identification prompt and its response displayed, click the three dots on the top right next to your prompt name, and select Compare.
Comparing by modifying system instructions
8.While compare Inside of the System instructions box (for the prompt on the right side) edit with below:
You are an expert insurance risk analyst assistant. Your task is to identify potential risk factors from a given scenario. For each risk factor, also briefly suggest a potential mitigation strategy or question for the underwriter. Be clear and structured.
9.Wait for both prompts to generate responses. Now you can see compared output.
10.Comparing with a different temperature setting:
11. System instructions box for the prompt on the right,revert it to:
You are an insurance risk analyst assistant. Your task is to identify potential risk factors from a given scenario. Be concise.
12.Right side Model settings : Temperature to 2.0.
13.Right side Submit the following prompt in the box at the bottom of the page:
Scenario:
"The applicant, 'The Fiery Grill,' is a new upscale restaurant specializing in wood-fired oven pizzas and open-flame grilling. They have installed a brand new, custom-built fire suppression system for their cooking area, but it has not yet been certified by a third party. The restaurant plans to feature live acoustic music on weekend evenings and has a small, raised stage area. They also want to offer valet parking."
Based on this scenario, list three primary risk factors an underwriter should consider.
14. Observe the differences in the responses. Does the higher temperature (2.0) vs compared to the 0.2 temperature output?
15. Ensure the Model is set to model_name
16. Temperature: 0.2
17. Set the Thinking level to Minimal
18. Change following prompt(more complex) in the box at the bottom of the page.
Scenario:
"The applicant, 'The Fiery Grill,' is an upscale restaurant specializing in wood-fired ovens and open-flame grilling. They have a brand new, custom-built fire suppression system, but it has not been certified by a third party. The restaurant will feature live acoustic music on a small, raised stage. They also want to offer valet parking, managed by their own staff. The applicant has no prior business history."
Underwriting Guidelines:
Priority Hierarchy: Liability risks are classed as:
- Class A (Critical): Fire, structural failure, failure of safety systems.
- Class B (Standard): General premises liability (e.g., slip-and-fall).
- Class C (Niche): Auto/Vehicle liability.
Compounding Factors: A "compounding risk" (a condition that makes another risk worse) must be elevated to the highest priority.
Inexperience: Lack of prior business history is a general negative factor but does not create a primary risk on its own.
Auto Liability: Class C risks (Valet) are only considered a primary risk if the applicant is using an unvetted, third-party contractor.
Task:
Based on the scenario and the underwriting guidelines, identify the single, #1 highest-priority risk. Then, write a 2-sentence justification that explains why it is the #1 risk, citing the specific guideline(s) that apply.
19. Review the responses. Compare the output from model_name (left pane) with model_name (right pane).
20. Click Check my Progress.
Task 4. Use multimodal prompts with Gemini
1.analyze an image and extract information from it.
2. analytical tasks such as description, text extraction, and question
3. Navigate back to the Agent Studio page.
4.On the top left corner, click New Chat.
5.Change the name of the prompt to Timetable Image Analysis.
6.On the bottom left of the prompt box, click the + button, then select Import from Cloud Storage.
7.In the menu, select the pre-built Cloud Storage bucket, then the timetable.png file.
8.In the Model settings panel on the right:
9.Ensure the model_name model is selected.
10.Ensure the Region is Global.
11.Now, you can ask the model to perform a few tasks on the image. Below the inserted image in the prompt input field, paste the following prompt:
1. Provide a concise title for this image (under 5 words).
2. Describe the image in one or two sentences.
3. Extract all visible text from the image. Present the flight schedule as a clearly formatted list with columns for "Time" and "City".
12. Submit/Enter
13.Next, you can ask a question that requires reasoning based on the extracted information. Submit the following prompt:
Based on the flight schedule shown in the image, what percentage of the listed flights depart before 11:30 AM? Show your calculation if possible.
14. Submit/Enter
15. Setting: Temprature 0.8
16. Check output and again Tmprature set:0.2 to get more predictable response. Can be set 0 to 0.2 lower for getting good factual response.
17.Radio button Check my progress click and check this task 4 completed.
Task 5. Generate media in Agent Studio
Generate an image with Gemini
1.In the left navigation menu, under Generate media, click Image.
2.In the text prompt area at the bottom, enter a descriptive prompt:
A close-up, photorealistic image of a single honeybee collecting pollen from a vibrant purple lavender flower, with a softly blurred garden background.
3.In the Settings panel on the right:
3a.Ensure the Model settings is set to model_name.
3b.For Aspect ratio, select 1:1.
3c.Review the Safety Filter Settings (e.g., Person generation, Safety filter threshold) and leave them at their defaults or adjust as preferred for this non-person-focused prompt.
4. Submit/Save
5.Now Image will be generated.
6.Click the generated image thumbnails to open the detail view.
7.Click Radio button Check my progress to verify the objectives.
Task 6: it is Optional related to - Generating Voice with Chirp
If you wish to explore AI-generated voice:
1. In the small vertical toolbar on top, click the Speech icon (it looks like a microphone) to switch to the voice generation tool. This opens the Chirp interface.
2. You might be prompted to enable the Cloud Text-to-Speech API if it's not already active for your project. If so, click Enable and wait for the API to be enabled (this might take a moment).
3. Once the interface is ready, in the text prompt area at the bottom, enter the text you want to synthesize. For example:
Welcome to the world of generative AI on Google Cloud
4.In the Settings panel on the right:
4a.Select a Model (e.g., Chirp 3. HD Voices).
4b.Choose your desired Language (e.g., English (United States)).
4c.Select a Voice from the dropdown list. You can try a few different ones to hear their characteristics.
4d.Explore any Advanced options if available and desired.
5.Click the Run button.
6. After processing, you should be able to play the generated audio directly in the interface.
Congratulations! In this lab, you successfully navigated Agent Studio
Comments (0)
No comments yet. Be the first to share your thoughts!