Enterprise · Deployment
Cloud Private Cloud and On Premise Deployment
Run Hamsa in our cloud, inside your own cloud account, or on your own infrastructure. Choose by where your data must live, which systems your agents must reach, and who will run the platform.
- Hamsa Cloud
- Private cloud in your VPC
- On-premise on your infrastructure
- Language model set per agent
Options
Hosted, private cloud or self-hosted voice AI
The same models and APIs in each option. Private cloud and on-premise are part of the Enterprise plan.
Hamsa Cloud
Hamsa hosts and runs the platform for you. Data is encrypted in transit and at rest, and your team builds agents in the browser without setting up any infrastructure.
Private cloud
The full Hamsa stack deployed inside your own VPC or dedicated tenancy. It runs in your cloud account, under your encryption keys and your network rules.
On-premise
Models, storage and logs run on your own infrastructure, in your own data center. For organisations whose policy keeps voice data inside the building.
Data flows
Where the data in a voice AI call goes
Before choosing where the platform runs, map what each part of a call sends and to whom.
Speech in, speech out
Hamsa’s own speech models turn the caller’s voice into text and the agent’s reply back into speech.
The language model
An agent’s replies come from the language model you set for it: OpenAI, Gemini, Groq, DeepMyst, or any OpenAI-compatible endpoint. Conversation text goes to the provider you choose, so choose one your data policy allows.
Your systems
Tools call your APIs during the call. Webhooks send call events and outcomes to an HTTPS endpoint you control.
Recordings and transcripts
Calls are recorded and transcribed for review in call history. Recordings are reached through signed URLs, and access to call history is set by role.
Comparison
Cloud, private cloud and on-premise compared
Private cloud and on-premise are on the Enterprise plan. Contact us to scope either one.
| Option | What it means |
|---|---|
| Hamsa Cloud | Hosted and run by Hamsa; encrypted in transit and at rest; no infrastructure on your side |
| Private cloud | Full Hamsa stack in your VPC or dedicated tenancy; your cloud account, encryption keys and network rules |
| On-premise | Models, storage and logs on your own infrastructure |
| Models and APIs | The same in every option |
| Language model | Set per agent: OpenAI, Gemini, Groq, DeepMyst or any OpenAI-compatible endpoint; Gemini by default |
| Plan | Private cloud and on-premise are on the Enterprise plan |
Decision guide
Choosing between cloud, private cloud and on-premise voice AI
Four questions to answer with your security and infrastructure teams. The answers point to an option.
- 01
Where must the data live?
Write down what your policy or regulator says about call audio, transcripts and recordings. If a provider may hold them, Hamsa Cloud is the shortest path. If they must stay in your cloud account or your building, look at private cloud or on-premise.
- 02
Which language model may see conversation text?
Agents send conversation text to the language model you configure. Decide which providers your policy accepts, or whether you will use an OpenAI-compatible endpoint of your own.
- 03
Which systems must agents reach?
Agents call your APIs through tools and take calls on your Twilio numbers or SIP trunk. List those connections and the firewall rules they need.
- 04
Who will run it?
Hamsa runs Hamsa Cloud. With private cloud or on-premise, the cloud account or hardware, the keys and the network are yours, so name the team that owns them.
FAQ
Questions about voice AI deployment
Can Hamsa run on-premise?
+
Yes. The Enterprise plan includes on-premise and private cloud deployment with the same models and APIs as Hamsa Cloud. On-premise, the models, storage and logs run on your own infrastructure.
What is the difference between private cloud and on-premise?
+
Private cloud deploys the full Hamsa stack inside your own VPC or dedicated tenancy, under your cloud account, encryption keys and network rules. On-premise runs it on your own infrastructure, in your own data center. Both are on the Enterprise plan.
Which plans include private cloud and on-premise?
+
The Enterprise plan, which also includes unlimited usage and custom terms. Contact our team to scope either option for your environment.
Which language model do the agents use?
+
The one you choose for each agent: OpenAI, Gemini, Groq, DeepMyst, or any OpenAI-compatible endpoint. Gemini is the default. Flow agents can use a different model for each node. Conversation text is sent to the provider you configure, which matters when you decide where data may go.
Is data encrypted in Hamsa Cloud?
+
Yes. Hamsa Cloud encrypts data in transit and at rest. Credentials for your systems are stored as encrypted project secrets, and webhooks are delivered only to HTTPS endpoints.
How do agents connect to our phone lines and systems?
+
Phone calls come through numbers you hold in Twilio or through your own SIP trunk. Agents reach your systems through tools that call your APIs or MCP servers, and send call events to your endpoints by webhook. Your firewall may need to allow Hamsa’s addresses for SIP.
Keep reading
Enterprise · Security
Security Compliance and Data Control
Isolated projects, roles, public and secret keys, encrypted secrets and HTTPS webhooks, with the language model set per agent and private deployment options.
Read →Enterprise · Services
Enterprise AI Implementation Services
Solution engineers take Arabic voice agents through discovery, flow design, integration and go-live, with support in your region and time zone.
Read →Hamsa Agents · Phone
AI Phone Agents
Inbound and outbound Arabic voice agents on your Twilio numbers or SIP trunk, with keypad input, transfers and campaigns.
Read →


