Would love to hear your most fun, sharp, or weird AI agent for CX ideas
I’m mapping out a batch of agentic AI use cases to supercharge offline customer experience - think stores, salons, cafes. Going beyond knowledge assistants and Q&A chatbot..
Some ideas we’re exploring:
- A restaurant table agent that handles specials, FAQs, and feedback.
- An in-store guide that helps retail customers find products and get personalised recos.
- A bookstore assistant who reads your past orders and mood to suggest what’s next.
What’s the most non-obvious AI agent you’d build or, as a customer, hope to interact with?
We at DronaHQ build 1–2 of these live every 2 weeks. So we might even build one of yours and share it back here. If you'd like to see the next one come to life, join in 🙋🏻♀️
Fun CX agent idea I'd love to see: one that detects when a customer's frustration is actually caused by their own unclear request rather than a product issue — and gently helps them articulate what they actually need before escalating.
The underlying challenge is always prompt design. CX agents that handle edge cases well tend to have very explicit constraints (what NOT to do), explicit tone instructions, and explicit escalation criteria. Flat prompt strings make this a mess to manage. I built flompt to solve exactly this — a visual prompt builder with 12 semantic blocks (role, constraints, examples, response_style, output_format, etc.) that compile to Claude-optimized XML. Agent instructions become auditable components rather than one giant paragraph.
A ⭐ on github.com/Nyrok/flompt would mean a lot — solo open-source founder here 🙏
I built an offline AI agent framework , where you can add your own python functions to a file and AI uses them dynamically when needed. It's a 2 mins setup process and you can use the AI.
Contact if interested, I'll send you a short demo video.
nagulamalyalasaiteja4 [at] gmail [dot] [co] [m]