
Pulastya AI
pulastya.sdlccorp.comDocument-based AI voice agents
Build AI voice agents that answer and make business calls using your own documents and voice choices.

Overview
Pulastya AI is a platform for building AI voice agents that answer and make phone calls for your business. You choose the voice, upload your documents, set the instructions, test the agent in your browser and follow every call from one dashboard. It is for businesses that want to automate inbound and outbound calls with human handoff when needed.
How it works
When a customer calls your Twilio or Exotel number, or your agent places an outbound call, the caller's words are transcribed as they speak. Real-time OpenAI voice streaming runs on Twilio. The agent follows your instructions, answers from the PDFs you upload, and replies in the voice you picked or transfers the call to your team. Every call is logged with status, transcript, summary, intent and sentiment.
Key capabilities
Choose voice providers including Twilio Polly (Amazon Polly or Google TTS), OpenAI TTS and ElevenLabs, with Indian, British and US English voices. Upload PDFs to Knowledge for the agent to answer from, and use prompt templates for Healthcare, E-commerce, Banking, Real Estate, Hospitality, Education or a general agent. The dashboard lists every call with totals, filters by status and date range, search by phone number, and export the call list. Transferred calls have their own status, filter and count.
Who it is for
Pulastya AI is for businesses that want to automate phone calls, including customer support, healthcare clinics, e-commerce, real estate, hospitality and education. Each industry comes with a starter prompt, and you can add your own documents and adjust instructions. It is designed for teams that need to hand off calls to humans when necessary, with role-based access for users, managers, admins and super admins.
Pricing and limits
Pricing depends on your use case and call volume. Plans include Starter, Professional and Enterprise, all with custom pricing on request. Every plan runs on your own Twilio or Exotel account and your own OpenAI API key. The Professional plan adds OpenAI TTS and ElevenLabs voices, real-time OpenAI voice streaming on Twilio, human handoff with transferred-call tracking, call summaries, intent and sentiment, call list export, and dashboard filters.
Key features
Inbound and outbound AI calling
Answer incoming calls and place outgoing calls through Twilio or Exotel, with real-time OpenAI voice streaming on Twilio.
Knowledge base from PDFs
Upload FAQs, policies and service guides; the agent answers callers from these documents and tells them when information isn't covered.
Prompt templates for industries
Start with ready-made prompts for Healthcare, E-commerce, Banking, Real Estate, Hospitality, Education, or a general agent, then edit to match your business.
Human handoff
Transfer calls to your team with full context; transferred calls get their own status, filter and count on the dashboard.
Call analytics and transcripts
Every call is logged with status, transcript, summary, caller intent and sentiment; filter by status and date range, search by phone number, and export the call list.
Browser testing
Test the agent from your browser with your microphone and a live transcript, or check answers by text in Test AI Chat before going live.
Use cases
Customer support
Answer common support questions from help documents and pass complex issues to your support team.
Healthcare
Handle appointment and clinic enquiries from clinic FAQs, and hand over to staff when a patient needs a person.
E-commerce
Handle routine service questions, review transcripts afterwards, and transfer sensitive requests to staff.
Real estate
Respond to property enquiries from listings and process documents, and connect interested buyers with an agent.
Hospitality
Take booking and guest enquiries about rooms, check-in, check-out, amenities and reservations, day or night.
Education
Answer admissions and course questions from course documents and hand callers to the admissions team when needed.
Pricing & limits
Verified 10/29/2026Frequently asked questions about Pulastya AI
- What is Pulastya AI?
- Pulastya AI is a platform for building AI voice agents that answer and make phone calls for your business. You choose the voice, upload your documents, set the instructions, test the agent in your browser and follow every call from one dashboard.
- Which telephony providers does Pulastya support?
- Twilio is the default. Exotel is supported for standard call flows: outbound connect, connect-to-flow, inbound webhook, status callbacks and transfers. Real-time OpenAI voice streaming currently runs on Twilio only.
- Do I need my own API keys?
- Yes. Pulastya works with your own accounts: you add your Twilio or Exotel credentials and your OpenAI API key. They are stored encrypted and masked in the workspace.
- Can a call be handed to a person?
- Yes. The agent can transfer a call to your team. Transferred calls get their own status, filter and count on the dashboard.
- Can I test the agent before going live?
- Yes. Start a test call in your browser to talk to the agent with your microphone and see a live transcript, or check its answers by text in Test AI Chat.
Related tools

Ramsa Speech
Turn text into natural Emirati Arabic speech for live conversations and studio-quality voice-over.

Lyria 3 AI Music Generator — Create Songs from Text & Lyrics Free
Turn text prompts or your own lyrics into complete songs with vocals and instrumentals in seconds.

Elmren Voice
Turn documents into local speech, audiobooks and read-along files on Mac

GenSong
Turn a prompt or lyrics into a complete song with vocals and instrumentals in minutes.

Palabra.ai
Translate live voice and audio in 60+ languages with sub-second latency and voice cloning.

Text to Song
Turn text or lyrics into complete songs with vocals in under 30 seconds, no musical experience needed.

Inwista
Transcribe video and audio in 70+ languages, identify speakers, and export subtitles in SRT, VTT, or MP4.

Transcribe Audio to Text
Turn audio, video, and YouTube links into editable transcripts with timestamps and speaker labels.