Siermet Descriptive Data Analysis Agent

Siermet needs an intelligent agent that focuses on data analysis—not sales or general customer-support triage—so every feature you build should revolve around extracting, summarising, and presenting insights from our datasets. The core capability is descriptive analysis: spotting patterns, highlighting trends, and explaining “what happened” in the clearest possible way.

Even though the main role is analytical, I want users across all our support channels to access those insights seamlessly. That means the same agent must be able to respond via live chat, craft email replies, and, where feasible, supply concise talking points for our phone team. Think of it as an analysis engine that can surface answers wherever our audience happens to ask.

Please rely on proven tools you’re comfortable with (Python, Pandas, SQL, or equivalent stacks) and design the interaction layer so it can plug into standard live-chat widgets, IMAP/SMTP for email, and a straightforward API our phone system can hit.

Deliverables:
• A deployable agent that performs accurate descriptive analytics on our existing data sources
• Connectors or webhooks for live chat, email, and phone hand-off
• Clear setup documentation plus a quick demo dataset so I can validate results locally

Acceptance criteria will be a short recorded walkthrough that shows the agent answering the same analytical question across all three channels with consistent, correct output.