The assumption is half right: sell LiteLLM Enterprise by running a technical, buyer-led evaluation, not a feature pitch

Asked (summary):

Preparing a 30-minute mock discovery call as a prospective LiteLLM enterprise AE, with a Head of AI Platform Engineering evaluating litellm.ai. Told LiteLLM does not do value-based selling given the product — how will the call go? Is it just answering questions and selling the benefits of Enterprise?

This playbook is built from 24 sourced findings across 5 independent hosts — LiteLLM's product, Enterprise docs, and pricing pages, current AE role listings, and one independent production-practitioner write-up — checked as of 14 Sep 2026. It gives a minute-by-minute plan, a question bank, an OSS-vs-Enterprise map, and a scorecard.

Verdict: partly right. "Different discovery" means architecture-led and evidence-led — not discovery-free. Since the buyer can get an instant 30-day Enterprise trial with no sales call, the AE's job is not to gate access or dump features: it is to diagnose fit (scale, governance gaps, deployment, support needs), map only confirmed requirements to the OSS/Enterprise boundary, and make the trial decisive. LiteLLM's own qualification bar: 100+ users or 10+ production AI use cases.

The 30-minute call, as one flow

Minute-by-minute plan with sample phrasing

Question bank for a Head of AI Platform Engineering

Hypothesis-led, ordered roughly as the call runs. Ask a handful well; do not interrogate.

OSS versus Enterprise: where the line actually sits

The core gateway is the same MIT-licensed codebase. OSS covers the fundamentals; Enterprise is for scaled organizational governance, deployment options, support, and procurement. Never imply gateway basics are paywalled.

Capability area
Open source
Enterprise adds

Likely technical questions — answer, don't bluff

Rule: clarify the question, answer what the docs support, and book an SE/engineering follow-up for anything deeper. Overclaiming to this buyer ends the evaluation.

Buyer asks
Suggested response
Source

What strong and weak look like in the mock

Strong

  • Concise technical fluency: speaks proxy, K8s, SSO/SCIM, virtual keys natively
  • Hypothesis-led questions built on what the buyer already runs
  • Listens; lets the buyer's architecture drive the agenda
  • Precise qualification against the 100+ users / 10+ use-case bar
  • Honest boundaries: "I'll confirm with engineering" beats a bluff
  • Maps only confirmed needs to Enterprise features
  • Closes on a concrete mutual action plan with trial success criteria

Weak

  • Generic ROI theater and value-based interrogation for pain
  • Reading every Enterprise feature off the docs page
  • Overclaiming capabilities or roadmap (e.g. the Rust migration)
  • Letting the whole call become free support / Q&A
  • Implying OSS is crippled to force the upgrade
  • Closing without success criteria or a named next step

Scripts: opening, transition, close

Opening (minute 0)

"Thanks for the time — I know you can grab a 30-day Enterprise trial key yourselves without talking to us, so my goal today isn't to gate anything. I'd like to understand how you route LLM traffic now, see whether Enterprise is even relevant at your scale, and if it is, leave with a trial plan your team actually cares about. Sound fair? Anything you want to make sure we cover?"

Mid-call transition (around minute 17)

"Given what you've described — self-hosted OSS proxy, roughly forty teams on virtual keys, SSO past the 5-user free tier, and a multi-region requirement — three things you named sit on the Enterprise side: SSO at your scale, audit logs, and multi-region under one license. The routing and fallbacks you asked about are already in the open-source core, so you don't buy those. Want me to walk through just those three?"

Close (minute 28)

"Here's what I suggest: I email the 30-day trial key today, you validate SSO with your IdP and the admin/worker multi-region split in staging, and we set the bar now — say, p99 overhead and a passing security review. I'll bring an engineer to a 45-minute working session next week and loop in whoever owns procurement, since AWS or Azure Marketplace may let you use existing commit. Does Thursday work?"

Interviewer scorecard

Dimension
1–5
What a 5 looks like

All 24 sourced findings

CategoryFindingCall implicationSource

Method: 24 sourced findings, one per factual consideration, drawn from 5 independent hosts (official LiteLLM product, Enterprise docs and pricing pages, current AE role listings, and one independent production-practitioner article), checked as of 2026-09-14. The most-cited single URL (LiteLLM Enterprise docs) backs 6 of 24 rows (25%). Call-plan timings, scripts and scorecard are coaching guidance built on those findings; each factual claim links its source. Nothing was cut; all 24 rows appear in the evidence table.

This report was generated automatically by Keenable SELECT at a user's request, from publicly available web sources linked herein. Keenable does not review, verify, or endorse its contents and makes no representation as to accuracy, completeness, or timeliness; AI-based extraction may contain errors. Nothing in this report is investment, legal, financial, or other professional advice. All trademarks and referenced content remain the property of their respective owners; no affiliation or endorsement is implied. To report an error, rights concern, or request removal: legal@keenable.ai.

Keenable SELECTAsk your own question
Made with Keenable SELECT