AskSenseOps Prerequisites
Table of Contents
- Server Requirements
- Software and Database Requirements
- Network Requirements
- BI Platform Requirements
- LLM Provider Requirements
- SSL Certificate Requirements
- MCP Client Requirements
- MCP Access and Governance
This section guides you through the prerequisites required to install and configure AskSenseOps. The requirements vary depending on whether AskSenseOps is connected to Power BI, Qlik Sense Enterprise, or Qlik Cloud.
Server Requirements
AskSenseOps requires a dedicated server with the following configuration:
| Requirement | Minimum |
|---|---|
| Operating system | Windows Server 2019 or later, Ubuntu 20.04 or later, or RHEL 8 or later |
| CPU | 4 cores |
| Memory | 8 GB RAM |
| Storage | 50 GB available |
| Network | 1 Gbps with low latency to the target BI environment |
Recommended configuration:
- 8 or more CPU cores
- 16 GB RAM
- 100 GB or more SSD storage
- Linux operating system
The reference configuration supports approximately 10–20 concurrent users.
Software and Database Requirements
AskSenseOps requires the following software on the server:
| Software | Requirement |
|---|---|
| Python | 3.12 or later |
| Node.js | 18 or later |
| PostgreSQL | 14 or later |
| pgvector | Required PostgreSQL extension |
Create an empty PostgreSQL database for AskSenseOps. The database user must have permission to create tables, indexes, and extensions.
Default database name: senseops_chatbot
Default PostgreSQL port: 5432
Network Requirements
Ensure that the following ports are available on the AskSenseOps server:
| Port | Purpose |
|---|---|
| 3010 | AskSenseOps |
| 443 | Secure access (HTTPS) |
The AskSenseOps server must have internet access through HTTPS (port 443) to connect to the selected BI platform and LLM provider.
If your environment uses a proxy or reverse proxy, ensure that it is configured to allow AskSenseOps traffic.
BI Platform Requirements
The required prerequisites vary depending on the BI platform used with AskSenseOps. Select your BI platform and complete the listed requirements.
Power BI
Microsoft Entra ID
Register an application in Microsoft Entra ID and keep the following details available:
- Tenant ID
- Application (client) ID
- Client secret or certificate
Grant the application the following Power BI API permissions:
Dataset.Read.AllReport.Read.AllWorkspace.Read.All
Administrator consent is required for these permissions.
If delegated authentication is used, configure the AskSenseOps sign-in URL as the redirect URI.
Power BI Access
- Enable the Power BI tenant setting that allows applications or service principals to use Power BI APIs.
- Give the AskSenseOps application and/or pilot users access to the required workspaces and datasets.
- Configure Row-Level Security (RLS) if different users need access to different data.
Power BI Network Access
| Host | Purpose |
|---|---|
login.microsoftonline.com |
Microsoft Entra ID authentication |
api.powerbi.com |
Power BI REST API |
*.analysis.windows.net |
Dataset queries |
*.powerbi.com |
Power BI service |
If the XMLA endpoint is used, the required Premium Per User, Premium, or Fabric capacity must be available for the datasets in scope.
Qlik Sense Enterprise
Complete the following requirements if AskSenseOps is configured to work with Qlik Sense Enterprise on Windows.
- Qlik Sense Enterprise February 2020 or later.
- A service account with Repository API access.
Qlik Sense Certificates
Export the following certificates from the Qlik Management Console (QMC):
client.pemclient_key.pemroot.pem
Export the certificates for the AskSenseOps server.
Virtual Proxy
- Add the AskSenseOps origin to the Qlik Sense virtual proxy allowlist.
- Keep the Qlik Hub URL and virtual proxy prefix available.
- Keep the authentication method available.
Supported authentication methods:
- Header
- Ticket
- SAML
- JWT
Ensure that the required streams and applications are available to the pilot users.
Network Access
| Port | Purpose |
|---|---|
| 4242 | Qlik Repository API |
| 4747 | Qlik Engine |
| 4243 | Qlik Proxy/Hub |
For multi-node environments, port 4747 must be accessible to the Engine node hosting the applications used by AskSenseOps.
User Access
- Pilot users must exist in the Qlik user directory.
- Pilot users must have access to the required streams and applications.
- Pilot users should be signed in to Qlik Hub when authorizing the connector.
- Third-party cookies must be allowed for the Qlik virtual proxy during authorization.
Qlik Cloud
Complete the following requirements if AskSenseOps is configured to work with Qlik Cloud.
Qlik Cloud Tenant
- Qlik Cloud tenant URL.
- OAuth M2M client ID.
- OAuth M2M client secret.
user_defaultscope.- Impersonation enabled for the OAuth client.
Web Integration
- Create a Web Integration ID.
- Add the AskSenseOps origin to the Web Integration allowlist.
- Ensure that the configured origin matches the AskSenseOps URL.
Spaces and Applications
- Required spaces and applications must be accessible to the pilot users.
- Pilot users must have at least Viewer access.
- Pilot users must have an appropriate Qlik Cloud entitlement, such as Professional or Analyzer.
Network Access
| Host | Port | Purpose |
|---|---|---|
<tenant>.<region>.qlikcloud.com |
443 | Qlik Cloud services and OAuth |
| Qlik Cloud tenant over WebSocket | 443 | Qlik Engine |
*.qlikcloud.com |
443 | Qlik Cloud resources |
If a custom Qlik Cloud domain is used, add the custom hostname to the outbound allowlist.
No inbound access to Qlik Cloud is required.
User Access
- Pilot users must exist in the Qlik Cloud tenant.
- Pilot users must have access to the required spaces and applications.
- Pilot users should be signed in to Qlik Cloud when authorizing the connector.
- Third-party cookies must be allowed for the Qlik Cloud tenant during authorization.
LLM Provider
Select an LLM provider and keep the required connection details available.
| Provider | Required details |
|---|---|
| Azure OpenAI | Endpoint, API key, deployment name, API version, model name |
| OpenAI, Anthropic | API key, model name |
| AWS Bedrock | Region, access key ID, secret access key, model ID |
| Google Vertex AI | Project ID, location, service-account credentials |
| Groq, Mistral, Cohere, Together, Fireworks, GitHub Models | API key, model name, base URL |
Allow outbound HTTPS access to the selected LLM provider.
| Provider | Host |
|---|---|
| Azure OpenAI | *.openai.azure.com |
| OpenAI | api.openai.com |
| Anthropic | api.anthropic.com |
| AWS Bedrock | *.amazonaws.com |
| Google Vertex AI / Gemini | *.googleapis.com |
SSL Certificate Requirements
A publicly trusted SSL certificate is required for the AskSenseOps MCP endpoint.
- Certificate in PFX / PKCS#12 format
- Certificate password
- Complete certificate chain
If the intermediate certificates are provided separately, store them in:
data/proxy/ca_bundle.pem
The minimum supported TLS version is TLS 1.2.
MCP Client Requirements
AskSenseOps can be accessed through the following MCP clients:
- Claude
- ChatGPT
- Cursor
- Microsoft 365 Copilot
Ensure that pilot users have the required subscription or license for the selected MCP client.
| MCP Client | Requirement |
|---|---|
| Claude | Web or desktop access |
| ChatGPT | Developer Mode |
| Cursor | IDE or web access |
| Microsoft 365 Copilot | Required Copilot Studio licensing and configuration |
MCP Access and Governance
Users connect to the AskSenseOps MCP endpoint using:
https://<your-senseops-host>/mcp
OAuth is used for authentication.
For controlled access, the following options can be configured:
- Admin approval
- User allowlist
Once all the prerequisites listed above are met, AskSenseOps is ready for installation and configuration. Ensure that the required access, credentials, and network connectivity are available before proceeding.