AI solutions we build
- 01WhatsApp AI assistants (lead qualification, booking, support)
- 02RAG chatbots trained on your documents and knowledge base
- 03AI agents that act inside your systems (CRM updates, bookings, follow-ups)
- 04AI features in existing apps (summaries, classification, search)
- 05Guardrails, human handoff and evaluation, so the AI doesn't make up answers
Tech stack
- Python/FastAPI
- Node.js
- LangGraph
- Google Gemini and other LLM APIs
- Qdrant vector DB
- PostgreSQL
- Redis
- Twilio/WhatsApp Business API
How an AI project runs
- 01Use-case workshop
- 02Data and knowledge-base prep
- 03Prototype in week 1–2
- 04Guardrails and testing
- 05Launch on WhatsApp or web
- 06Monitor and improve
Key takeaways
- Starter. ₹35k–1.2 L excl. GST, 2–3 weeks. FAQ/WhatsApp chatbot or one AI feature in an existing app.
- Growth. ₹1.5–4 L excl. GST, 4–8 weeks. LLM/RAG bot on your documents, guardrails, admin, analytics.
- Advanced. From ₹4.5 L (custom quote) excl. GST, 8–12 weeks. AI agents acting in your systems (CRM/booking), multi-channel, evaluation.
- Prices exclude GST and LLM/WhatsApp API usage fees.
Plans and prices
| Plan | Price (excl. GST) | Timeline | Includes |
|---|---|---|---|
| Starter | ₹35k–1.2 L | 2–3 weeks | FAQ/WhatsApp chatbot or one AI feature in an existing app |
| Growth | ₹1.5–4 L | 4–8 weeks | LLM/RAG bot on your documents, guardrails, admin, analytics |
| Advanced | From ₹4.5 L (custom quote) | 8–12 weeks | AI agents acting in your systems (CRM/booking), multi-channel, evaluation |
Prices exclude GST and LLM/WhatsApp API usage fees.
See the AI development service.
Get a free AI consultation.
FAQ
At FYT, ₹35k–1.2 lakh for a FAQ or WhatsApp bot, ₹1.5–4 lakh for a RAG bot on your documents, and from ₹4.5 lakh for AI agents, plus GST and API usage.
Yes. VetOsHealth.ai and Rajwada's assistant both run on WhatsApp.
It answers only from your knowledge base (RAG), falls back safely when it doesn't know, and hands urgent cases to a human.
We pick per use case (e.g. Google Gemini in VetOsHealth.ai) and keep the design model-agnostic.
Your data stays in your database and vector store, and access is role-based.
Yes, that's the Starter plan.



