Agents
Delegate the work, keep the judgement.
Agents that read your sources, use your tools, and show their work — so your team reviews decisions instead of assembling context.
Agents
7 activeToday
Support triage — inbox #general
Retrieving·38 tickets
Vendor MSA review — Nordwind
Comparing clauses·64 clauses
RFP security questionnaire
Drafting gaps·112 / 118
Weekly engineering digest
Complete·5 themes
This week
Onboarding FAQ refresh
42 answers updated
Contract renewal watchlist
9 accounts flagged
Help-centre gap analysis
17 missing articles
Meet your team where they work
The same agents, the same knowledge, wherever the question is asked.
Web app
A workspace for chat, sources, agents and evaluation in one place.
import { Ragenta } from "ragenta"
const r = await agent.ask(
"refund window?"
)
API & SDK
Mint a workspace key and start an agent run from your own product.
Embedded widget
One script tag puts a grounded assistant on your own site.
Ragenta
3 contracts renew this month — Nordwind is the one to look at first.
Slack
Digests, alerts and answers posted into the channel that needs them.
Start
begin
Read the inbox
agent · gmail_search
Sort by topic
categorize
Draft a reply
llm
Open an issue
agent · github_create_issue
Post to Slack
agent · slack_post
Draft with a prompt
Read yesterday's email, sort it by topic, and handle each kind.
Nothing is saved until you publish a version.
Build an agent you can actually look at
Steps, branches and loops on a canvas instead of one long prompt. Describe the job in a sentence and Ragenta drafts the flow for you to correct — and every version you publish is kept, so a change can be undone.
See how the canvas works →It understands your organisation, not just your documents
Context, conventions and permissions carry into every run.
Contract terms
done · opus-4.5
Account history
running · gpt-5
Support tickets
running · gpt-5
Billing records
queued · qwen3
Parallel subagents
A broad question fans out across subagents, each answering one part and streaming findings back as it finishes.
“What did we commit to Nordwind on pricing?”
Knowledge with provenance
Every passage carries its source and where in the document it came from, and every run records what each step was given — so an answer can be audited, not just read.
Never quote an unapproved price
all agents
Cite the contract, not the summary
legal, sales
Escalate below 0.8 groundedness
support
Reply in the customer's language
support
Rules your team writes once
Tone, mandatory disclaimers, forbidden sources, escalation thresholds — written once and applied to every future run.
The whole loop, in one system
Ingest, retrieve, evaluate — and the evaluation feeds back into the first two.
Parse — layout-aware, tables preserved
Chunk — on structure, not character count
Embed — multilingual, one vector space
Sync — incremental, only what changed
Ingest
Connect a source and Ragenta parses layout, chunks on structure, embeds and keeps it in sync — with the source's permissions attached.
Retrieve
Vector similarity and keyword matching run together, then a reranker orders what actually reaches the model's context.
score 30 reference questions
→ groundedness 0.94 (+0.06)
check citation coverage
→ 98% of claims linked
diff against last configuration
→ 2 answers changed
Evaluate
Groundedness, citation coverage and retrieval hit rate are scored on every change, so a regression surfaces before a user reports it.
What an agent can actually do
Read, act, and leave a trail you can follow.
$ retrieving across 4 sources…
reranked 38 → 6 passages
$ sql: revenue by region, last 4 quarters
answer drafted · 9 claims · 9 citations
Run tools, not just prompts
Search your knowledge, query a database read-only, read a web page, work through a spreadsheet, send mail — with credentials held server-side and never shown to the model.
Point at exactly what matters
Mention a document, a folder, a connector or a saved filter inline, and the agent scopes its retrieval to it for that run.
Knowledge base re-indexed
24 Aug · 09:12
Reranker enabled
24 Aug · 11:40
Model switched to opus-4.5
25 Aug · 08:05
Current configuration
26 Aug · 14:22
Every run can be replayed
Model, prompt, tools and each step's input and output are recorded against the version that produced them, so any answer can be read back exactly as it was produced — or compared against another version on the same question.
Put it on your own website in one line
A chat bubble that answers from your knowledge base, on your domain. The allowed origins are checked on our server, every conversation is billed to your workspace, and you can read the questions visitors actually asked.
See what the widget reports →Do you deliver to Da Nang, and how long does it take?
Yes — Da Nang is next-day for orders placed before 4pm, and two days after that. Delivery is free above 500,000 ₫.
2 sources
Paste it before the closing body tag
<script src="https://app.ragenta.cloud/widget.js"
data-key="wk_live_8f2c…"></script>128
Conversations today
94%
Answered from your documents
3,410
Credits spent
Extend it without forking it
Templates to start from, connectors for the systems you already use, and your own tools alongside ours.
Connectors
Connect an account once, over OAuth, and its tools become selectable on any agent step — reading what the connecting person can read, and nothing else.
Research agent
Searches, reads and cites before it answers
Customer support agent
Answers from your policies, escalates when unsure
Data analyst agent
Asks your database, read-only, and shows the query
HR assistant
Answers internal policy questions with the source
Sales assistant
Drafts replies grounded in approved material
Document agent
Reads a document and pulls out what you asked for
Six agents to start from
Each template arrives as a working flow with its steps, tools and instructions already set — including what it must not do. Change what you need and publish a version of your own.
Your own tools
Attach any Model Context Protocol server, or register a plain HTTP tool, and it appears to your agents beside the built-in ones — with every call and its result recorded.
From the teams running it every day
“We had three separate search tools and none of them could answer a question. Ragenta replaced all three in a month, and the citations are the reason people actually trust it.”
Head of Operations
Logistics, 900 staff
“Support deflection went from a slide in a strategy deck to a number on a dashboard. The retrieval trace is what let us debug it ourselves instead of filing tickets.”
Customer Support Lead
B2B SaaS
“Permission-aware retrieval was the whole deal for us. Legal would not sign off on anything that indexed HR files into one bucket.”
IT Director
Financial services
“Our documentation is half Vietnamese, half English. One index handles both and an English question still finds the Vietnamese paragraph.”
Engineering Manager
Fintech, Ho Chi Minh City
“The agent does not just retrieve — it decides what to look up next. Watching it re-query after a weak first result was the moment this clicked for the team.”
Head of Data
Healthcare group
“We swapped the underlying model twice in a quarter without touching a single prompt. That flexibility is worth more than any single model benchmark.”
Principal Engineer
Enterprise software