An AI-accelerated 30-60-90 is a ramp plan where the slowest parts of getting productive — absorbing product knowledge, learning the market's language, and practicing under pressure — run through AI instead of waiting on other people's calendars. Traditional ramp takes as long as it does because it is mostly queuing: for enablement sessions, for shadowing slots, for the mistakes that teach. AI removes the queue.
Why Ramp Is Slow (It's Not the Content)
New reps are not short on information — they drown in it. Wikis, decks, recorded calls, tribal knowledge scattered across channels. What they lack is retrieval and reps: the ability to get an answer the moment a question forms, and safe at-bats before the real ones. Both are exactly what AI supplies.
Days 1–30: Build Your Second Brain
Before anything else, build a persistent AI workspace — a Claude Project or equivalent — loaded with the product docs, pricing, ICP, top case studies, and your five best recorded call transcripts. Then interrogate it daily:
Prompt: "Quiz me on [product area]: ask me ten questions a skeptical buyer would ask, one at a time. After each answer, correct me using the loaded docs and show me how a top rep on the recorded calls phrased it."
Every question you would have saved for your manager's Thursday 1:1 now gets answered in seconds — with the source. Thirty days of that compounds into what used to take ninety.
Days 31–60: Practice Before You Play
Month two is at-bats. Run daily AI role-plays against your market's actual personas — built from those loaded transcripts, so the objections are the real ones. Sequence the difficulty: friendly buyer week one, skeptic week two, hostile committee by week four. Get scored after every session, and keep a running list of the responses that still wobble. By the time you take live calls, nothing a buyer says is happening to you for the first time.
Days 61–90: Sell With a Safety Net
Month three is live selling with AI as the debrief partner: every call transcript goes through a post-mortem prompt — what did I miss, what MEDDIC data is still open, what does the follow-up need to say — and every deliverable (emails, recaps, discovery plans) gets drafted against the workspace before it ships. You are selling at week nine with the context depth of a second-year rep, because the context was never the bottleneck — access to it was.
For Managers Reading This
The rep builds the workspace; you seed it. The five transcripts you choose become the new rep's model of "good." Choose calls that show recovery and discovery craft, not just smooth wins — reps learn more from watching a save than a parade.
Frequently Asked Questions
Can AI really cut ramp time in half?
The queue-bound parts, yes — knowledge retrieval and practice reps compress dramatically. Relationship-building and territory trust still take calendar time; the win is that they start sooner because the rep is dangerous earlier.
What should go into a new rep's AI workspace first?
Pricing, ICP, objection responses, two competitor one-pagers, and five great call transcripts. Resist loading everything — a curated workspace answers better than an archived one.
Does this replace enablement and shadowing?
No — it removes the waiting between them. Shadowing teaches judgment; AI makes sure the hours between shadow sessions build skill instead of anxiety.
Put It to Work
Promptifi's library includes onboarding quiz prompts, role-play scenarios, and call-debrief workflows a new rep can run from day one. Browse the library.