Yes. The bot tracks conversation state across children, so a parent enquiring for two kids on different STEM tracks gets two matched tutors with two separate trial bookings. It applies family-pack pricing if your packs allow mixing across subjects and tutors. Most STEM services have multi-kid families as a common pattern, and the bot's structured intake reduces the back-and-forth that usually happens.

Yes, if your tutor profiles list competition experience (USACO Gold or Platinum, AMC 10 honor roll, FRC programming, ARML). The bot recognises the competition by name and routes to a tutor who has actually competed. Most generalist tutors are not the right match for serious competition prep; the bot's job is to filter for the specialist. For services without competition specialists, the prompt should say so and route those questions to a referral list.

Yes. If your roster includes university-level tutors (research methods, advanced calculus, algorithms, organic chemistry, biochem), the bot reads those profiles and routes university students accordingly. Pricing at the university level is usually higher and the conversation pattern is different (one-off help on a specific problem set or longer support across a semester). The bot adapts when university course codes appear in the question.

Yes, if your packs are on the pricing page. The bot quotes single-subject hourly, multi-subject family packs, and competition-prep packages from real data. It will not invent packs; if you want it to honour a referral or sibling discount that is not on the pricing page, add the rule to the prompt. Otherwise it quotes only what is published.

Yes, if your roster includes project-based tutors (FRC, FTC, VEX, Arduino, Raspberry Pi, science fair coaching). The bot reads those profiles and routes project-based questions accordingly. Project-based help is often a different conversation pattern (longer arc, deliverable-driven) and pricing usually reflects that. Most services run a separate landing page for project work, and the bot adapts its tone and recommendations when the visitor enters from there.

Late April is brutal for STEM services. AP Calc, AP Physics, AP Chem, AP Bio, AP CS A all converge in two weeks. The bot scales because it reads from WordPress and calls your model API. There is no admin bottleneck. Conversation logs let you see which APs are spiking before they peak, so you can open new sections or shift tutor hours rather than turning students away. This is one of the strongest ROI cases for STEM services.

Yes. Hand off to TutorBird, Calendly, Acuity, or any tool with a public booking URL via deep link or webhook. The bot collects subject, level, and competition context per child in chat, then passes the family to your booking link or fires a webhook. For multi-kid bookings, most services use webhooks because deep links handle single bookings better than batched ones.

Logs stay in your WordPress database. Bring-your-own-key sends prompts directly to OpenAI, Anthropic, or Google with no third-party processor in between. That matters for students sharing grades and struggles: you control retention and redaction, and the plugin does not add a new vendor between you and the model. Apply your existing WordPress security to the log table.