Case Study / Agentic AI · Customer Support
Deployed forSMAC Tech LabsAI customer support with human handoff for SMAC Tech Labs
Algorims built SMAC Support on Amazon Bedrock and Claude to answer questions from the company's knowledge base and hand complex cases to agents with the conversation history intact.
- Client
- SMAC Tech Labs
- Industry
- Digital Engineering & IT Services
- Platform
- Amazon Bedrock (Claude)
- Status
- Live in production
Measured impact
Live in production, handling real customer conversations
AI agent grounded in SMAC Tech Labs' own knowledge base
Seamless AI-to-human handoff with full conversation context
01 / Overview
The operating context.
SMAC Tech Labs — a digital engineering company delivering cloud, mobile, web, and AI/ML services out of Singapore and India — needed a support experience for its own customers that matched the AI-native systems it builds for its clients.
Algorims designed and built SMAC Support end-to-end: a dedicated AI support platform where an AI agent answers customer questions directly from SMAC Tech Labs' own knowledge base, and hands off to a human agent the moment it can't help.
02 / Challenge
What stood in the way.
SMAC Tech Labs needed a support channel that could scale without scaling headcount, while staying grounded in accurate, SMAC-specific answers:
- 01
Customer questions needed answers grounded in SMAC Tech Labs' own knowledge base — not generic AI responses that risk being wrong.
- 02
Support needed to be available continuously, without a growing team watching every channel manually.
- 03
When the AI couldn't help, the handoff to a human agent needed to be seamless, with conversation history carried over rather than lost.
- 04
The platform needed to be built specifically around SMAC Tech Labs' own workflow, not a generic third-party chatbot bolted on top.
03 / Solution
What Algorims built
SMAC Support — a purpose-built AI support platform on Amazon Bedrock, combining a knowledge-grounded AI agent, an embeddable chat widget, and a full agent dashboard.
Knowledge-grounded AI agent
An AI agent built on Amazon Bedrock (Claude) answers customer questions from SMAC Tech Labs' own knowledge base rather than general training data, so answers stay accurate and on-brand.
Embeddable web chat widget
A lightweight, embeddable chat widget puts the AI agent directly on SMAC Tech Labs' site, answering questions in real time without a support agent needing to be online.
Seamless human handoff
When a question needs a human, the conversation escalates to SMAC Tech Labs' support team with the full history intact — customers never have to repeat themselves.
Agent dashboard
A dedicated dashboard gives SMAC Tech Labs' team a live inbox, ticketing, and knowledge-base management, built specifically around how their support team works.
System stack
Production foundations.
- Amazon Bedrock
- Claude (Anthropic)
- NestJS
- Next.js
- PostgreSQL
04 / Results
What changed after delivery.
Live in production at SMAC Tech Labs' support platform, handling real customer conversations today.
AI-first, human-backed — questions are answered instantly by the AI agent where possible, with a clean handoff to a human for anything it can't resolve.
Purpose-built for SMAC Tech Labs — a dedicated platform reflecting their own knowledge base and support workflow, not a shared generic tool.