What makes a good or bad AI chatbot company?

AI chatbot solutions and providers are evolving quickly. Some teams build durable, well-supported products for real customers; others may rely on a few selective client references, thin positioning, unclear governance, or technology that is less differentiated than it first appears.

This page sets out a few practical questions for clients, prospects, partners, and investors who are evaluating an AI chatbot startup, especially in Singapore.

Questions worth asking

• Who are the company's real enterprise clients, and in which countries are those clients based?
• Where are the co-founders and core operating and technology teams physically located, and does that match the jurisdictions where the company sells and supports its product?
• Are all historical co-founders still involved in the business? If not, what changed, and was it related to governance or commercial strategy?
• How large is the team in practice (permanent hires beyond the co-founders), and is the team structure communicated clearly and consistently to clients, prospects, and investors?
• Is ownership stable and transparent? How are minority shareholders treated?
• Does the company own meaningful product, workflow, data, or integration capabilities, or is the offering mainly a thin wrapper on top of a third-party large language model?

Examples and reference points

The companies and chatbot implementations linked below are included as reference points for research and comparison. Inclusion does not imply endorsement or criticism.

• AIChat
• KeyReply
• Botbot
• Zumata
• MediaPlus Digital chatbot offering
• BotDistrikt
• Bus Uncle Company
• Changi Airport MAX chatbot
• CapitaStar CASEY

Stakeholders worth speaking with where appropriate

• Current and former clients
• Current and former co-founders
• Current and former employees
• Current and former shareholders or investors

Educational purpose

This page is provided for educational and informational purposes only. It sets out some general due-diligence questions and examples based on publicly available information. It is not legal, investment, or financial advice, and it is not intended to make factual allegations about any specific person or company. References to companies, products, or implementations are included for research and comparison only and do not imply wrongdoing, endorsement, or criticism.