Google Workflow for Invoice Image Processing -- 2
Budget: $70 – $100 USD
I'm looking for a talented developer who can create a Google Workflow that will get an image from Google Cloud Storage, perform base64-encoding on that image and then initiate Gemini Pro Vision (Multiple Model).
**This is USD 100 fixed price project.
Required Capabilities:
- Knowledge of Google Workflows in REST API
- Google Cloud Platform.
- POSTMAN
- Familiarity with Gemini Pro Vision
This workflow is necessary for optimal compatibility with Gemini Pro Vision, a tool I plan to use for visual Q&A. The questions to be answered will primarily involve identifying specific text and details from invoice images. If you have experience in this specific type of visual Q&A, your skills would be highly desirable for this project. Your efforts will lead to increased efficiency in organizing and extracting information from our invoices.
I look forward to hearing from individuals with the right expertise. Your involvement in this project will vastly streamline our current invoice handling procedures.
***
Google Cloud Platform Account is ready
After the configuration,
you need to have a short Zoom meeting to provide the guideline for the configuration to the other GCP account.
This is long term project, there are still many tasks for google workflows are ready for you.
===========
Details
===========
1. Send an Image (Invoicing) to Google Cloud Storage from POSTMAN
* Using Http REST API end point
* 0Auth2.0 is not really yet
* Public Access to the Photo on Google Cloud Storage
* Save photo into the "Invoice" Folder.
* Save Respond in POSTMAN
* Image will be provided
2. Send Image URL from POSTMAN to Google Workflow with custom field values.
*Get Image URL from Step 1.
*Custom Field values: (value will provided)
1. NSID
2. Email
3. Phone Number
3. POSTMAN Get Respond from Google Workflow
*Postman will received "Success" Respond
*Postman will STOP working from here.
*Google Workflow will continue start working.
4. Generate Image to Based64
from Google Cloud Storage into Google Workflows
*save the Code for next step in workflows
5. Send Based64 Code to Gemini Pro Vision (Multiple Model)
specific type of visual Q&A and to generate output.
* Prompt will provide to you
* Insert Based64 code with Prompt
* Insert required Parameter into API.
* The Output is Json Format
* Save the output for next step.
References
* https://cloud.google.com/vertex-ai/docs/generative-ai/model-reference/gemini
* https://cloud.google.com/vertex-ai/docs/generative-ai/multimodal/send-multimodal-prompts
6. Send The Json Output to MongoDB Atlas in google workflows
* API Endpoint and Parameter will provide to you.
* Save the Respond from MongoDB.
7. Final Step
Send the MongoDB Respond via Webhook in Google Workflows
* Webhook URL will be given
* Payload contain MongoDB Respond and Custom Field from Step 1
**This is USD 100 fixed price project.
Required Capabilities:
- Knowledge of Google Workflows in REST API
- Google Cloud Platform.
- POSTMAN
- Familiarity with Gemini Pro Vision
This workflow is necessary for optimal compatibility with Gemini Pro Vision, a tool I plan to use for visual Q&A. The questions to be answered will primarily involve identifying specific text and details from invoice images. If you have experience in this specific type of visual Q&A, your skills would be highly desirable for this project. Your efforts will lead to increased efficiency in organizing and extracting information from our invoices.
I look forward to hearing from individuals with the right expertise. Your involvement in this project will vastly streamline our current invoice handling procedures.
***
Google Cloud Platform Account is ready
After the configuration,
you need to have a short Zoom meeting to provide the guideline for the configuration to the other GCP account.
This is long term project, there are still many tasks for google workflows are ready for you.
===========
Details
===========
1. Send an Image (Invoicing) to Google Cloud Storage from POSTMAN
* Using Http REST API end point
* 0Auth2.0 is not really yet
* Public Access to the Photo on Google Cloud Storage
* Save photo into the "Invoice" Folder.
* Save Respond in POSTMAN
* Image will be provided
2. Send Image URL from POSTMAN to Google Workflow with custom field values.
*Get Image URL from Step 1.
*Custom Field values: (value will provided)
1. NSID
2. Email
3. Phone Number
3. POSTMAN Get Respond from Google Workflow
*Postman will received "Success" Respond
*Postman will STOP working from here.
*Google Workflow will continue start working.
4. Generate Image to Based64
from Google Cloud Storage into Google Workflows
*save the Code for next step in workflows
5. Send Based64 Code to Gemini Pro Vision (Multiple Model)
specific type of visual Q&A and to generate output.
* Prompt will provide to you
* Insert Based64 code with Prompt
* Insert required Parameter into API.
* The Output is Json Format
* Save the output for next step.
References
* https://cloud.google.com/vertex-ai/docs/generative-ai/model-reference/gemini
* https://cloud.google.com/vertex-ai/docs/generative-ai/multimodal/send-multimodal-prompts
6. Send The Json Output to MongoDB Atlas in google workflows
* API Endpoint and Parameter will provide to you.
* Save the Respond from MongoDB.
7. Final Step
Send the MongoDB Respond via Webhook in Google Workflows
* Webhook URL will be given
* Payload contain MongoDB Respond and Custom Field from Step 1