Build vs buy: the eighteen-month tail
Internal AI study-tool builds typically reach a working prototype in four to six weeks. The team then discovers the eighteen-month operational tail over the following year: keeping up with Anthropic's quarterly Claude releases, regression-testing prompts across model upgrades, maintaining multilingual generation quality across nine languages, satisfying GDPR audit requirements, supporting open export formats (Anki, RemNote, Brainscape, Notion, OneNote, Kahoot) and handling PDF ingestion edge cases (scanned documents, unusual fonts, image-only pages, OCR fallback). StudyFlow ships against this entire surface with weekly releases. Customers get day-one operational maturity, under their own brand, at a fraction of a single ML engineer's loaded annual cost. Annual Starter is €590; annual Business is €1,990 — both for fewer euros than a single L&D engineer's monthly compensation in most European markets. The build-vs-buy comparison is rarely close once the eighteen-month tail is priced honestly.
Not an LMS — the AI study layer beside it
StudyFlow is not a Learning Management System. It is the AI study layer that produces what an LMS does not — chapter summaries, Bloom's-tagged exam questions, exam-date-aware study plans and spaced-repetition mini-exams from the institution's own source material. Generic LMS platforms (Moodle, Canvas, Cornerstone OnDemand, Docebo, SAP SuccessFactors, 360Learning, Brightspace, Blackboard) deliver content and manage enrolment; trainees who study seriously end up in ChatGPT alongside the LMS — invisible to the institution. Most StudyFlow customers deploy alongside an existing LMS rather than replacing it. The LMS continues to handle enrolment, content delivery and reporting; StudyFlow handles the trainee study workflow. SSO via Google and Microsoft / Azure means trainees move between the two with one set of credentials. No integration build is required for the initial deployment, and tenant data stays isolated per institution.
What ships under your brand on day one
- Chapter detection and structured summaries. Every chapter is split into load-bearing main topics and supporting side topics, with key terms surfaced as a distinct field.
- Bloom's-tagged exam and discussion questions. Up to 8 exam + 5 discussion questions per chapter, weighted toward apply / analyse / evaluate. Multiple-choice distractors are grounded in side-topic content, not fabricated.
- Exam-date-aware study plan. Realistic daily allocations keyed to the trainee's stated availability.
- Spaced-repetition mini-exams. Post-session quizzes interleave prior chapters — spaced retrieval rather than massed practice.
- Exam-readiness dashboard. Per-chapter readiness scores plus admin aggregates on the Business tier.
- Open export formats. Anki, RemNote, Brainscape, Notion, OneNote, Kahoot — trainees keep their materials when their seat ends.
- Multilingual native generation. Nine languages, run directly on Anthropic Claude — not translated from English.
- Custom branding and domain. Logo, palette, product name, subdomain or apex domain — your brand on every screen.
Pedagogy as a primitive, not a prompt
Main-vs-side topic separation, Bloom's tagging, interleaved retrieval and SM-2-style spacing aligned to the study plan are first-class fields in the output schema — not behind a prompt that another team could replicate over a weekend. The structural commitments are visible to the institution and inspectable per-chapter, which matters for buyers whose data-protection officer (or learning-design lead) wants to see how generation behaves rather than trust a vendor's "AI-powered" claim.
Start a Starter tenant in under five minutes
Self-service provisioning. €59/month for 25 trainee seats (Starter) or €199/month for 100 seats (Business). Pay via Stripe; branded space live in under five minutes — no sales call, no setup fee, cancel anytime. Annual billing at the equivalent of ten monthly payments — two months free on both tiers.