Products we build in-house.
Three products that work together and are tailored for each client: an AI agent that answers and acts, an inbox where AI and your CS team work side by side, and a gateway that connects everything to AI models.
Otora AI Agent
A conversational agent that answers customers from your business knowledge and takes action in your systems: checking stock, creating orders, sending payment links, and tracking deliveries.
- Knowledge base from your catalogue, SOPs, FAQs, and documents.
- Actions in your systems through connections to ERP, online stores, and payment gateways.
- Many channels: WhatsApp, Telegram, Instagram, and web chat.
- Clear limits: important actions wait for your team’s approval.
Hi, is the navy linen shirt still available in size L?
Yes, we have 4 in our Samarinda warehouse. Shall I create the order for you?
Yes please, send it to the same address as last time.
Order SO-2231 is ready, total IDR 289,000. You can pay with this QRIS link.
Omnichannel inbox
One screen for every customer conversation. The AI agent handles routine questions, your CS team handles what needs a person, and the full history stays in one place.
- WhatsApp, Telegram, Instagram, e-mail, and web chat in one inbox.
- Assign conversations to team members and track their status.
- An automatic summary of every conversation handed over by the AI agent.
- Customer history across channels and team performance reports.
- Rina · WhatsAppPaid already, thank you!AI agent
- Budi · TelegramWhen will my parcel reach Balikpapan?AI agent
- Santi · InstagramCan I order 50 portions for an office event?To CS team
- Visitor · Web chatDoes the ERP work for three branches?AI agent
Otora AI Gateway
A gateway that sits between your applications and AI model providers. One OpenAI-compatible API for GPT, Claude, Gemini, Llama, and local models, with routing, fallback, quotas, and cost tracking.
- One API, many modelsCompatible with the OpenAI SDK. Switch models by changing one parameter.
- Smart routingChoose a model per task by quality, speed, or cost.
- Automatic fallbackWhen a provider is slow or down, requests move to a backup model.
- Keys & quotas per teamUsage limits and budgets for every app, team, or client.
- Visible costsTokens and spend per model and per app, in rupiah.
- Data controlMasking of sensitive data, a no-log option, and local models for data that must stay in-house.
Access to Otora AI Gateway is currently available to project clients and partners. Contact us for a demo and to discuss your needs.
from openai import OpenAI
client = OpenAI(
base_url="https://api.otora.id/v1",
api_key=OTORA_API_KEY,
)
reply = client.chat.completions.create(
model="auto", # the router picks a model
messages=[{"role": "user",
"content": "Summarise this customer complaint"}],
)| Task | Primary model | Fallback |
|---|---|---|
| Replying to customer chats | Claude | Gemini |
| Summarising long documents | Gemini | GPT |
| Sorting tickets and e-mail | Llama (local) | Mistral |
AI connected to every platform you use.
AI models in the centre, customer channels and internal systems around them. One gateway keeps everything consistent, secure, and easy to change.
- Commercial and open-source models, including self-hosted ones.
- Official channels: WhatsApp Business Platform, Telegram Bot API, Instagram, and web chat.
- Internal systems via API: ERP, CRM, warehouse, and payment gateways.
Tell us which work you want to hand to AI.
Send a short description by WhatsApp or e-mail. We reply with follow-up questions and a suggested first step.
- +62 853-4817-3761Phone and WhatsApp
- fikri@otora.idE-mail
- Samarinda, Kalimantan TimurRegistered office
