Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
hoverbot
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
1.
▲
by
hoverbot
1y ago
We just built a production grade multilingual RAG system and wrote up the design and what we learned in “Multilingual RAG That Actually Works: A Practical Architecture.” Many multilingual RAG setups break in familiar ways: embedding spaces
2.
▲
Close the loop: analytics that teach your chatbot to fix itself
(hoverbot.ai)
8 points
by
hoverbot
1y ago
|
4 comments
3.
▲
Four chatbot widget patterns for websites and apps: from bubble to super app
(hoverbot.ai)
2 points
by
hoverbot
1y ago
|
1 comments
4.
▲
by
hoverbot
1y ago
We outlined four common patterns for embedding chatbots, from simple bubbles to multi-tab hubs, and looked at where each approach fits best. Would love to hear what patterns the HN community has seen work well (or fail) in real-world use.
5.
▲
Make a smart router the center of your chatbot
(hoverbot.ai)
2 points
by
hoverbot
1y ago
|
1 comments
6.
▲
by
hoverbot
1y ago
Author here. We run chatbots with a smart router: classify noise/PII first, call retrieval+LLM only when needed, and use deterministic flows for actions. This keeps latency and cost down and makes behavior explainable. What would you a
7.
▲
Handling PII in customer-facing AI chatbots: mask before sending to LLM
(hoverbot.ai)
2 points
by
hoverbot
1y ago
|
1 comments
8.
▲
by
hoverbot
1y ago
Author here (HoverBot team). How we keep PII out of LLMs in our chatbots: detect fast -> replace with typed placeholders -> call the model -> unmask on our server. We also cover region pinning and vendor privacy controls. Keen to h
9.
▲
by
hoverbot
1y ago
Great practical takes - agentic chatbots work only with real data and feedback loops. HoverBot is built this way: it automates RAG pipelines, includes confidence thresholds, and supports human-in-loop override. You can run vertical assistan
10.
▲
by
hoverbot
1y ago
Tough experience, but it doesn’t have to be. With HoverBot, if the bot is unsure, it knows when to back off and bring in a human right away. And with built-in conversation analytics, you can keep fine tuning your flows so things just keep g