Build guide

How to build a customer support assistant with Vikasit AI

A support assistant answers customer questions from your help docs and knows when to escalate to a human. Ground it in your knowledge base with retrieval so it stays accurate and on-brand.

Recommended model

Vikasit 3

A dependable generalist for grounded, on-tone support answers at low cost. Use Vikasit 3 Max for complex, multi-step troubleshooting.

Steps

1
Index your help docs and FAQs into a vector store for retrieval.
2
For each customer message, retrieve the most relevant articles.
3
Build a prompt with the articles, the conversation history, and tone guidelines.
4
Instruct the model to answer from the docs and to escalate when unsure.
5
Detect escalation signals and hand off to a human agent when triggered.
6
Log conversations to improve docs and measure deflection rate.

Code

The Vikasit Inference API is OpenAI-compatible, so this uses the standard OpenAI Python SDK pointed at https://api.vikasit.ai/v1.

customer-support.py

from openai import OpenAI

client = OpenAI(
    base_url="https://api.vikasit.ai/v1",
    api_key="sk-vikasit-...",  # get one at vikasit.ai/auth
)

def support_reply(question: str, kb_articles: list[str]) -> str:
    kb = "\n\n".join(kb_articles)
    resp = client.chat.completions.create(
        model="vikasit-3",
        messages=[
            {
                "role": "system",
                "content": (
                    "You are a friendly support agent. Answer from the knowledge base. "
                    "If you can't answer confidently, reply EXACTLY 'ESCALATE'.\n\n"
                    f"Knowledge base:\n{kb}"
                ),
            },
            {"role": "user", "content": question},
        ],
    )
    return resp.choices[0].message.content

Build your customer support assistant today

Get an API key and 2M free tokens a day on Vikasit Nova. Pay-as-you-go, no minimums, OpenAI-compatible.

Get an API key See all models & pricing