1. How quickly can a customer get up and running with Pi Agent?
Pi Agent, also known as the "multi-agent" due to its multi-agent capabilities, is deployed on Customer's Google Cloud Platform. The process involves three steps:
- Installation from the GCP Marketplace.
- Configuration, integrating with the customer's source data.
- Insights generation, enabling end-user access from Pi agent UI.
This initial setup can typically be completed within a few days, depending on the complexity of the data integration. It would take up to 4 weeks (white glove professional service) for limited use cases and limited data sources.
2. How does Pi Agent integrate with BigQuery?
Pi Agent comes with Planning in a Box for data foundation to host structured and unstructured data in customer's GCP tenant and is part of the Google Gemini Enterprise. Structured data is typically stored in BigQuery, while unstructured data leverages GCS. All these components are accessible through a single interface with the Pi Agent VM within the same GCP region.
3. How does the pricing model work?
The pricing model consists of two components:
- A licensing fee for Pi Agent, provided by Pluto7 through the Marketplace.
- Consumption charges for the underlying Google Cloud resources used, including (but not limited to) the Compute Engine VM hosting Pi Agent, BigQuery, Vertex AI for model training, and related services. GCP consumption is additional and can vary from $5K to $8K per month depending on customer.
The pricing is consumption-based, calculated on a per-hour basis. Self-service options are available with limited support, while enhanced professional service support is available upon request through Pluto7. Contact support@pluto7.com for support requests.
4. What if a customer needs professional services?
Pluto7 offers "white glove" professional services for Pi Agent. These services cover configuration, integration, onboarding, and ongoing support, enabling monthly releases and continuous value realization.
5. What are some similar products in the market?
Customers often compare Planning in a Box Pi Agent with generic enterprise AI platforms like C3.ai and Palantir. However, Pi Agent is purpose-built for specific verticals by use cases E.g. Inventory, Forecasting, Manufacturing Solution Accelerators etc, which is well suited for various size businesses.
6. What is the underlying architecture?
Pi Agent's architecture is based on the standard provided by Google Cloud engineers as best practices. Capabilities are enabled as needed. Pi Agent is designed to work out of the box with minimal configuration to get started. Pluto7 strongly recommends customers to engage with Pluto7's professional services for optimal and enterprise level implementation. Reach out to support@pluto7.com.
7. What kind of data sources does Pi Agent support?
Pi Agent supports a wide range of data sources for integration. However, each source must be evaluated to ensure effective integration. Pluto7 recommends professional services, including discovery sessions, to bring in domain expertise and tailor the solution to the customer's specific needs.
8. What is the typical timeframe for configuration?
For customers with sample data, a basic setup can be completed in one day and configured for user review within two weeks for users to login and check the platform with initial sample data. For larger enterprises with more complex projects, Pluto7 recommends a four-week MVP with one use case and, a four-week implementation, and a six-month enhancement and maintenance plan, including support for proper value extraction.
9. How does licensing work?
Licensing follows Pluto7's platform usage model and is based on adoption. The license is not for resale or distribution. Contact Pluto7 for further details at support@pluto7.com.
10. How do end-users access and utilize the output?
End-users access the output through the customer's own Google Cloud tenant.
11. Where does all the work, data logic, and code run?
All processing occurs within the customer's Google Cloud tenant.
12. How do we engage with Pluto7?
13. Where can I find more information?
Visit www.pluto7.com for more information, including case studies, customer testimonials, and collateral.
14. What is the estimated monthly consumption cost?
The estimated monthly consumption cost for Google Cloud resources, excluding the cost of the VM itself, can vary significantly from $5,000 to $8,000 per month. This range depends on the size and scale of the business, data volume, and the complexity of the AI models used. This cost is in addition to the Pi Agent licensing fee and VM fees listed in the marketplace.
15. What documentation guide can one refer for Planning in a Box Pi Agent?
Please visit Gemini Enterprise, and refer to the "Documentation Guide" for more information.
16. What is the deployment model (cloud-based, on-premise, hybrid), and what are the infrastructure requirements (hardware, software, network)?
The solution is generally cloud-based, leveraging the Google Cloud Platform (GCP).
Infrastructure requirements include:
- Google BigQuery as the primary data warehouse for storing and processing structured data.
- Cloud Composer for data pipelines.
- Cloud Storage for file storage.
- Cloud Run for deploying and scaling containerized applications.
- Compute Engine instances may be used to host applications.
- The Cortex Framework is deployed within the customer's GCP tenant.
17. How does the solution integrate with existing systems (e.g., CRM, ERP, databases), and are APIs available?
- Solutions integrate with existing systems such as SAP ERP (ECC), SAP IBP, Salesforce, and Oracle, using APIs and data connectors.
- Real-time data replication from SAP ECC is achieved using Aecorsoft.
- A Master Ledger centralizes data from various sources, providing a unified view.
18. How does the system scale to handle increasing data volumes and user traffic, and what are the typical performance metrics (e.g., response times, throughput)?
- The system architecture supports horizontal scaling to handle increasing data volumes and user traffic.
- The solution should deliver real-time or near real-time responses to user queries and updates.
19. What security measures are in place to protect data, and does the solution comply with relevant industry regulations (e.g., GDPR, HIPAA)?
- Data is encrypted at rest and in transit, adhering to industry best practices and GCP's security protocols.
- Cloud IAM is used to manage roles and permissions, following the principle of least privilege access.
- Role-based access control (RBAC) is implemented to restrict access to sensitive data and functionalities based on user roles and responsibilities.
- Solutions are designed to comply with relevant industry standards and regulations like GDPR and CCPA.
20. To what extent can the solution be customized to meet specific business needs, and what tools and processes are available for customization?
- The solution can be customized to meet specific business needs.
- Tailored dashboards and reports can be created using tools like Google Data Studio, Looker, and SAP Analytics Cloud.
- Gemini Flash API can be utilized and prompts engineered for data extraction and transformation.
21. What level of support is provided (e.g., phone, email, online), and what are the maintenance procedures and schedules?
- Pluto7 provides technical support to address issues or queries during implementation and ongoing operation.
- The system is designed for straightforward updates and upgrades, ensuring compatibility with new versions of dependent technologies while minimizing downtime.
- Please Contact for more information support@pluto7.com
22. Where is the data stored, what are the data retention policies, and what are the storage capacity limits?
- Structure data stored in BigQuery where unstructured data stored in Cloud Storage.
- Data is stored in BigQuery. BigQuery has generous storage capacity limits. Data retention policies depend on the dataset configuration, and can be customized based on requirements.
23. How are user roles and permissions managed, and can granular access control be implemented?
- Cloud IAM is used to manage roles and permissions, following the principle of least privilege access.
- Role-based access control (RBAC) is implemented to restrict access to sensitive data and functionalities based on user roles and responsibilities.
- Granular access control can be implemented.
24. What monitoring and logging capabilities are provided, and can system performance and usage be tracked?
- Comprehensive error handling and logging mechanisms are required to identify, log, and diagnose issues promptly.
- Detailed logs should be accessible for performance monitoring and troubleshooting.
- Tools for monitoring system health, performance, and resource utilization must be included.
- Custom dashboards and alerts monitor the performance and latency of the application, ensuring timely response to issues.
25. What disaster recovery plans are in place to ensure business continuity, and what is the Recovery Time Objective (RTO) and Recovery Point Objective (RPO)?
- Reliable backup and recovery processes must be in place to safeguard data against loss and facilitate recovery during system failures or disasters.
- High uptime must be achieved, incorporating redundancy and failover mechanisms to minimize service interruptions and support continuous business operations.
- The sources do not provide specific details on Recovery Time Objective (RTO) and Recovery Point Objective (RPO).
26. Does Pi Agent come with MDM(master data management) or a data foundation?
- Pi Agent has built in MDM(master data management) called Master Ledger which is part of Planning in a Box data foundation enabled with Cortex.
