Dashboard Agent
Generates source-cited dashboards, refreshes on schedule, and explains every chart in plain language.
Reports · Live refresh
Product · Four modes, five engines
One co-worker. Four
One co-worker. Four
ways to put it to work.
A Riyalabs co-worker isn't a single product – it's one of four interaction modes, picked to fit how your team already works. Ask, Watch, Embed, or Build.
§01 · Modes
Four ways to use a co-worker.
Pick the mode that matches how your team already works. Same engines underneath – different surface, different posture.
01 · Ask
Explore Ask
02 · Watch
Explore Watch
03 · Embed
Explore Embed
04 · Build
Explore Build
Ask. Pull answers.
Q&A and analysis your team can trust because every answer cites its source.
Example
"What changed in last week's customer health?" → 3 accounts flagged, with source rows from Salesforce + Zendesk.
Watch. Push alerts.
Your co-worker monitors approved data streams and surfaces what changed – with a recommended next step.
Example
EU ticket volume up 47% week-over-week. Recommended: alert CS lead + open priority queue. Awaiting approval.
Embed. Call from anywhere.
Slack, your CRM, ticketing, internal tools. The co-worker shows up inside the workflow – not in a separate tab.
Example
In Salesforce, hit "Brief me" on a deal – the co-worker drops a cited summary right into the record, ready for review.
Build. Ship a workflow.
Hire a brand-new co-worker for a specific workflow, or promote a successful conversation into a repeatable, approval-gated job.
Example
"Make this weekly. Same sources." → Workflow saved. Runs Mondays at 8am. Approval required before send.
§02 · Engines
Agent engines we specialize in.
Each co-worker is assembled from these underlying engines. Deep expertise lives here, not on the surface.
Data Analysis Agent
Answers business questions with citations. Identifies likely causes across multiple approved sources.
Cited · Multi-sourceSimulation Agent
Runs what-if scenarios on your data – pricing, capacity, forecast variance – with assumptions you can audit.
Scenarios · Assumptions visibleGoal-Based Agent
Tracks a defined business goal week-by-week, recommends the next step, and routes the action to the owner.
Goal-anchored · Action-ledAnomaly Detection Agent
Watches your data streams. Flags outliers and drift before they grow, with thresholds you can tune.
Always-on · Threshold-tuned
§03 · Assembly
How a co-worker is assembled.
A clear, short loop – not a platform rollout. Pick a mode, pick a workflow, assemble the engines you need, ship under approval.
01 · Pick a mode
Ask, Watch, Embed, Build
Decide how the co-worker should show up – pull, push, in-tool, or as a saved workflow.
02 · Pick a workflow
One repeatable job
Choose one high-value workflow your team already repeats every week.
03 · Assemble engines
From five specialties
Compose the right mix – Dashboard, Data Analysis, Simulation, Goal-Based, Anomaly Detection.
04 · Ship under approval
Human in the loop
Approval gate, source visibility, audit log, and a clear rollback plan, day one.
Get started
Bring one workflow your team repeats every week.
Riyalabs will help assess whether it's a good candidate for your first AI co-worker, what data it needs, what controls are required, and how to prove value safely.