Ampligo — the brain behind every music campaign.
From the first inbound email to the order handed off to Joseph: intake, analysis, proposal and sale — plus the learning loop that gets sharper from every campaign outcome. One operations brain, replacing the old spreadsheets.
- 0
- campaign verticals
- 0
- Joseph endpoints
- 0
- API routes
Why it exists
From manual chase to a governed flow
Ampligo replaces the pre-sales spreadsheets and the scattered email threads with a single, auditable pipeline.
Before — spreadsheets & inbox
- Inbound requests chased across email threads and re-typed into spreadsheets.
- Track and artist numbers looked up by hand, one platform tab at a time.
- Proposals rebuilt from scratch, pricing from memory and gut feel.
- Won deals handed to execution over chat, with no versioned source of truth.
With Ampligo
- Every email and submission lands as a structured Opportunity, ready to work.
- Enrichment from the music providers is pulled in and kept fresh automatically.
- Scenarios and pricing drafted on the memory of every past proposal — editable.
- A versioned, read-only handoff package flows to Joseph; outcomes flow back.
What it does
The pre-sales flow, end to end
Ampligo orchestrates everything that happens before a campaign goes live — and learns from what happens after.
Intake
Inbound emails and submissions become structured Opportunities — deterministic parsing first, AI second.
Analysis & enrichment
Track and artist intelligence from the music providers: the numbers, the history, the commercial moment.
Proposal
Scenarios and pricing, drafted on the memory of past proposals — always an editable draft.
Order
An accepted proposal becomes an order, then materializes into the campaigns to deliver.
Handoff → Joseph
A versioned, read-only package of approved objects, handed off to Joseph — the campaign executor.
Learning loop
Synthetic outcomes from Joseph flow back in and make the next proposal smarter.
Campaign verticals
Four services, one pipeline
The same pre-sales engine feeds every service line; each is executed downstream by Joseph.
Playlisting
Curator allocations & placements
Track Ads
Paid media — the save is the KPI
TikTok UGC
Content pipeline & burner pool
Playlist Growth
Monthly cycles & renewals
Under the hood
How track ↔ playlist matching works
Every playlisting suggestion is grounded in audio signals, not guesswork. Here's what feeds it — the exact scoring logic stays internal.
Genre
Detected from audio analysis and refined by the analyst's editorial call on the track; playlists inherit genre from their own sampled tracks.
Mood
The emotional tone picked up in the same audio pass, for the track and for every sampled playlist track.
Tempo (BPM)
Beats-per-minute measured from the waveform; a playlist gets a range from the tracks sampled from it.
Energy
A 0–100 score derived from the track's audio arousal, averaged across a playlist's sampled tracks.
Tracks get these signals from an audio analysis pass. Partner playlists are profiled the same way: a sample of their tracks is pulled from the playlist itself, run through the identical analysis, then averaged into a playlist profile — no manual playlist tagging.
Where the AI helps
One rule: it proposes, you decide
AI never acts on its own. It drafts, suggests and explains its confidence; a human confirms before anything counts.
Intake parsing
Reads inbound emails and submissions, extracts the fields, proposes the Opportunity — you confirm the match.
Proposal drafting
Suggests scenarios, playlist shortlists and pricing from proposal memory, as a draft you edit before it ships.
Assistant over your data
Answers questions across orders, clients and history in natural language, grounded in what's actually in the system.
Principles
How it behaves
Nothing ships without a human review
Every automation produces an editable draft. No email, placement decision or client-facing action ever leaves autonomously; low-confidence AI lands in a review queue.
Deterministic first, AI second
Rules run before the model. AI fills the gaps and explains its confidence — it never silently overrides a deterministic result.
Mock-vs-live connectors
Every external provider degrades to a clearly-badged mock when its key is absent. Going live is a credentials change, not a code change.
A clear boundary with Joseph
Ampligo owns pre-sales, proposal memory, the handoff layer and the learning loop. Live execution is Joseph's. The two talk through a versioned contract.
Architecture
Two systems in dialogue
Ampligo packages the handoff and learns from the outcomes; Joseph runs the live campaign. Built on a deterministic-first foundation with audited AI.
External world
Ampligo
Pre-sales brain
Intake
Email & submissions → Opportunity
Analysis & enrichment
Track / artist intelligence
Proposal
Scenarios, pricing, proposal memory
Order
Won deal → campaigns materialized
Joseph
Campaign execution
Live campaign
Status & curator activation
Placements
Verified & rolled up
Client updates
Sent to the client
Outcome report
Pushed back to Ampligo
Foundation
Security & privacy
How the data is protected
Client data and commercial history are sensitive. Access is checked server-side, AI runs locally and audited, and nothing outward-facing happens on its own.
Identity & access
Staff sign in through Clerk with role-scoped access; every server action re-checks the role. Client proposal links carry their own signed, expiring token.
Private by default
Secrets stay server-side and out of version control. Data lives behind the app — there are no public object URLs.
Local, audited AI
Inference runs locally; every AI call is logged with model, tokens and confidence. Nothing leaves the box to a third party to make a decision.
Traceable actions
Who did what, and when, is recorded — a native audit trail rather than an afterthought bolted on later.
Ready to dive in?
Ampligo is the team's internal console. Sign in to pick up where the workflow left off; the in-app manual then walks through every screen.