About ElevenLabs SuccessOS

Built for the Customer Success — Strategic — Western Europe application at ElevenLabs.

What this app demonstrates

This app is designed to show three things simultaneously: deep understanding of ElevenLabs' product suite, a working model of how enterprise CS operates at ElevenLabs, and genuine AI-native operating style — not just in the job but in how the application itself was built and presented.

Vanessa Piacente said it directly: “The people who have impressed me the most as we're hiring are people who are going above and beyond — they've taken the time to learn ElevenLabs, even if they don't use ElevenLabs day-to-day, then put a demo together and sent it.”

This app does exactly that — and goes one step further by using ElevenLabs' own API to generate voice briefings. The AI CSM that speaks is not a gimmick: it mirrors precisely what Vanessa described building internally.

AI feature map

Account Brief

claude-sonnet-4-6Extended thinking + Streaming

Extended thinking shows the reasoning chain — not just conclusions. Streaming means a CSM sees the brief word-by-word rather than waiting. Together they mirror how a senior CSM would actually think through an account before a meeting.

QBR Composer

claude-sonnet-4-6Streaming

QBRs have a fixed structure that Sonnet executes well. Streaming gives the CSM instant feedback and makes the generation feel alive rather than a black-box wait.

Expansion Signal Discovery

claude-haiku-4-5Tool use (server-orchestrated)

The agent queries a mock ElevenLabs use case library, ROI engine, and expansion signal database — then synthesises. Tool use is server-orchestrated for Vercel's 60s timeout reliability; in production it would be a full agent loop.

Adoption Playbook

claude-haiku-4-5Self-reviewing single-pass JSON

Haiku generates the full 30/60/90 playbook AND a VP-level self-review critique in one call. The self-review is the guard rail — it surfaces what a skeptical VP would push back on before the CSM sends it to a customer exec.

Voice Briefing

ElevenLabs Eleven Multilingual v2ElevenLabs API + Claude script generation

Claude writes the script. ElevenLabs speaks it. This is the meta feature — a Customer Success tool for ElevenLabs that uses ElevenLabs' own voice technology. It mirrors the AI CSM motion Vanessa described: pulling account context and surfacing insights proactively, now narrated.

How this maps to the ElevenLabs CS model

Two core metrics

NRR and New Product Expansion — both tracked in the Revenue Engine and per-account.

Three product surfaces

ElevenAgents, ElevenCreative, ElevenAPI — all modelled with distinct metrics and use cases.

Adoption journey

First Build → Production → Expanding → Champion → Strategic. Critical 30-60 day window modelled.

AI CSM motion

Pulling context, surfacing proactive ideas, expansion signals from trialling data — mirroring Vanessa's internal build.

DACH / WE market nuance

GDPR, EU AI Act, German procurement cycles, French compliance culture — all flagged in prompts and data.

Whitespace expansion

Accounts given product access before contract (ElevenLabs' model) — trialling signals drive the expansion matrix.

Data note: All account data is synthetic. Company names are real but all metrics, stakeholder names, and use cases are fictional. No confidential data was used. The consumption time series is procedurally generated. See the portfolio to explore the accounts, or start with Deutsche Telekom for the most complete example.