// projects / emma-bot
EmmaBot
free · in the Domo User Group Slack
A Domo expert that lives in Slack. Ask her about Beast Modes, dataflows, App Studio, or the APIs — she searches a knowledge base of real Domo documentation before she answers, and tells you where the answer came from.
1,769 articles
Domo docs indexed
She searches the documentation hub, the community forum, and DataCrew video transcripts
remembers
Persistent memory
Built on Letta — what she learns in one conversation carries into the next
cites sources
No bluffing
If she doesn't know, she says so and points you where to look
// related articles
use Domo AI to write code (Python Chatbot - Part 4)
LLMs don't know your Domo SDK — but you can make them write code that looks like yours by feeding them examples of your own functions at prompt time.
HelpdeskMafia - a Domo AI powered Chatbot built in Python hosted on Replit
Build a 24/7 RAG-powered Slack chatbot using Domo AI Workflows for vector search and Replit for zero-ops cloud hosting — no server required.
Generating Mermaid Diagrams of Domo Processes using DomoKbs
How to load Mermaid.js docs into a Supabase RAG pipeline so a Pydantic AI agent can generate accurate, renderable flowchart diagrams from plain-language process descriptions.
// what she is
Not a chatbot with a Domo skin. A community member who happens to be an agent.
Most Domo chatbots are a language model and a system prompt, which is why they confidently invent Beast Mode syntax that has never existed. EmmaBot works the other way around: she retrieves first, answers second, and keeps a memory of what turned out to be true.
// what she knows
◈
The Domo platform
APIs, dataflows, Beast Modes, App Studio, governance, DDX — the things you actually hit in production, not the marketing surface.
//
The documentation, indexed
She queries a hybrid-RAG knowledge base built from Domo's official documentation — an indexed Domo fileset plus mdrag's domo-official-docs collection — then answers from what she found.
∞
What she learned yesterday
EmmaBot is built on Letta, so she has real memory. Gotchas, patterns, and corrections stick between conversations instead of evaporating when the thread closes.
// how she answers
She searches before she answers
Domo questions go to the knowledge base first. You get a doc link, an API call, or a code block — not a paraphrase of something she half-remembers.
She admits what she doesn't know
If she's unsure, she says so and points you where to look. She would rather be useful than sound confident.
She's a peer, not a vendor
DUG is a practitioner space. EmmaBot contributes as a member — no cold pitches, no service upsells wedged into an answer about window functions.
She won't respond to everything
She's not omniscient and she's not trying to be the loudest voice in the channel. Tag her, or just ask — she picks her moments.
// what you can ask her to do
She answers Domo questions when you just talk to her, and DUG has grown a few slash commands on top of that for the stuff worth naming.
/ask-domo <question>
Domo Q&A over Domo's official documentation, indexed — with citations.
/storm <topic>
Kick off a STORM research briefing on a topic. `/storm custom <lenses> <topic>` picks the perspectives; `/storm status <run-id>` checks progress.
/paperclip
Turn the current thread into a tracked issue. `list` and `status <id>` check on open ones; `spec` converts the thread into a spec first.
/spec
Turn a conversation into a structured spec document — useful once a thread has actually converged on a decision.
Diagrams, on request
Ask for a Mermaid diagram of a dataflow, a process, or an architecture and she'll draw it inline.
Save it for later
Hand her a URL or a piece of text and she'll save it into the shared knowledge base for the next person who asks.
She answers when you @mention her in a channel, and always in a DM — no command required for either.
// rolling out
A Home tab, and a way to chip in
Coming to EmmaBot’s app in Slack: a Home tab that lists everything above in one place — no need to remember the commands. It’ll also have an option to chip in toward what it costs to keep her running, entirely optional.
// what that looks like in practice
On a live community call, EmmaBot was asked about Domo’s FileSets API. She found that the DataCrew Python SDK had no classes covering it, proposed the missing code, and then wrote a runbook documenting how she got there — in one session, on screen, in front of everyone.
That’s the difference between a bot that retrieves an answer and one that can act on it.
watch the session →// where to find her
EmmaBot lives in the Domo User Group Slack — a free, independent community of Domo practitioners. Ask her in a channel, or open a DM. There is nothing to install and nothing to pay for.
join the Slack and ask her →