CLIP™

The Opportunity

Revenue is hiding inside contracts no one can read fast enough.

Every licensing deal locks up eligibility rules, window formulas and territory rights inside dense legal paragraphs. Today those rules are re-read by hand — slowly, inconsistently, at great cost. CLIP™ makes them executable.

⚠️

Today — Without CLIP™

Manual · 2–3 weeks · Error-prone · Every quarter

Planners manually read 80–100 page legal agreements every quarter
Eligibility rules buried in nested clause sub-sections
Title matching done via spreadsheets — titles slip through
Rights confirmed across 6 dimensions: title × territory × language × media type × license × window
Fee calculations with tiered rate cards, MGs, and escalators
Near-miss titles explored ad-hoc — no systematic tooling
Amendment layers cause silent version drift
15% revenue leakage from missed opportunities

✨

Tomorrow — With CLIP™

AI + Human · Minutes · Auditable · Continuous

Engine 1 extracts every rule in minutes (OCR + Agentic RAG + LLM cluster)
Engine 2 evaluates 12K titles × all categories × all terms in <2 seconds
Six-dimensional eligibility checked automatically (title/territory/language/media/license/window)
Fees, commissions, MGs calculated per term year automatically
Make-It-Eligible playbook — systematic near-miss detection with ranked levers
Amendment layers version-controlled; re-evaluation is automatic
Revenue pushed to ERP/Salesforce — no lag between eligibility and booking
Full audit trail — every rule decision explainable

The Moat · What makes CLIP™ impossible to replicate with a single LLM

Why CLIP™ Wins

Two agentic engines. One intelligence loop.

CLIP™ is not one model wired to a PDF. It is two purpose-built agentic-RAG engines working in a closed loop: Engine 1 reads the contract, Engine 2 reads the catalog — and tells you how to create revenue, not just discover it.

Contract rules are fixed. Your catalog is growing. CLIP turns that gap into a forward-looking, multi-year revenue playbook — title by title, term year by term year.

Agentic RAG
engines in a loop

N × M

Rules × titles ×
term years

Revenue opportunities
left hidden

The two engines at a glance

60-second view

Engine 1

Contract Rule Extraction

Agentic RAG · specialist LLM cluster

📄→🤖→🧬→🎯

Contract → Agent → Specialists → Rules in product categories

Input: fixed (the contract is the contract)

⇄closed
loop

Engine 2

Catalog Aggregator & Insights

Agentic RAG · pattern analyzer · eligibility reasoner

🗂️→🔍→💡→📈

Catalog → Patterns → "Make-it-eligible" playbook → Multi-year dashboards

Input: growing (new titles, cast, theatrical every week)

🔬 Go deeper Inside both engines — agentic RAG architecture, specialist LLMs, pattern analyzer & the Make-It-Eligible playbook Expand Collapse ›

Engine 1

Contract Rule Extraction Engine

Agentic RAG + cluster of specialist LLMs — turns 80–150 page contracts into structured rules grouped by product category.

📚

Rules span the entire term

A window in Year 3 depends on a cap exhausted in Year 1 and a rate card amended in Year 5. The agent plans retrieval across years and reconciles them.

🤖

Agentic RAG — not a prompt

Planner → Retriever → Reader → Verifier loop. Re-reads pages, verifies every rule against the catalog, escalates edge cases. One-shot LLMs fail at clause ~40.

🧬

LLM cluster by segment

Specialists for Qualifier, Start Date, Caps, Rate Card, Amendment run in parallel. Segment router places every rule in the correct product category — the bucket planners sell from.

🎯

Thinks like a planner

Hunts footnotes, territory carve-outs in images, "notwithstanding" clauses three amendments deep. Multi-modal across text, tables and images.

Engine 1 · architecture

Planner → Retriever → Reader → Verifier, then a specialist cluster

The agent plans, retrieves, reads (multi-modal), and verifies. Five specialist LLMs extract in parallel. A segment router places every rule inside the correct product category.

Agent

Planner

Decides the next clause / segment / term year to extract

→

RAG

Retriever

Vector + symbolic search across pages, tables & images

→

Multi-modal

Reader

Parses text, tables and embedded images / schedules

→

Grounded

Verifier

Checks every rule against Engine 2's catalog reference list

Qualifier

AND/OR trees

Start Date

Date algebra

Caps

Table fluency

Rate Card

Pricing matrices

Amendment

Redline diffs

Segment router → rules placed inside each Product Category (Pay-TV Premiere, SVOD Exclusive, Free-TV Window…).

Engine 2

Catalog Aggregator & Insights Engine

Agentic RAG + pattern analyzer + eligibility reasoner — applies Engine 1's rules to every title, predicts what's eligible, and tells you how to make more of your catalog sellable.

🗂️

Aggregates rules × catalog at scale

Every active contract × product category × term year × every title in the catalog. The agent plans the evaluation — not a brute-force cross-join.

🔍

Pattern analyzer, not a filter

Clusters near-miss titles; spots recurring disqualifiers ("12% of your catalog fails only on US theatrical screens"); surfaces systemic revenue gaps invisible to any BI tool.

💡

"Make-It-Eligible" playbook

For each near-miss title, concrete ranked levers: boost theatrical to 300+ screens, attach a tier-A cast member, shift release date by 6 weeks. A revenue-strategy assistant, not a report.

📈

Multi-term-year visuals

Year 1 / Year 2 / Year N dashboards showing how eligibility evolves as release dates, amendments and catalog changes flow through the rules. Planners see the curve, not just today's number.

Engine 2 · architecture

Aggregator → Pattern Analyzer → Eligibility Reasoner → Insight Narrator

A different flavour of agentic RAG — not extracting rules, but reasoning across the entire rule × catalog × time cube to find revenue and prescribe how to unlock it.

Agent

Aggregator

Joins every active rule with every title, per term year

→

ML + LLM

Pattern Analyzer

Clusters near-misses, finds systemic disqualifiers

→

What-if

Eligibility Reasoner

Generates ranked levers to tip each near-miss into eligible

→

Visual

Insight Narrator

Renders multi-year dashboards & planner-ready recommendations

Outputs — the reason CLIP sells

Today's eligible slate — per deal, per category, per term year
12-month forecast — titles about to cross eligibility thresholds
Make-It-Eligible playbook per near-miss title (boost theatrical · add cast · shift date · expand territory)
Multi-year visual dashboards — Year 1 · Year 2 · … · Year N

Fixed rules × growing catalog = compounding value.

Every new title, every theatrical result, every cast attachment instantly re-runs against every live contract. CLIP gets smarter as the catalog grows — and so does the revenue it surfaces.

New theatrical release → eligibility re-scored in seconds

Cast attachment → re-qualifies for premium buckets

Release-date shift → new windows open across deals

Amendment signed → only affected rules re-propagate

Future-title forecast → 12 months of pipeline visibility

Near-miss titles → ranked levers to tip them eligible

Pattern alerts → "12% of catalog fails only on screen count"

Multi-year view → revenue curve, not a snapshot

Platform Architecture

Two Agentic Engines. One Intelligence Loop.

Engine 1 — Contract Rule Extraction turns legalese into structured, executable rules. Engine 2 — Catalog Aggregator & Insights applies those rules to the catalog, predicts patterns, and tells you how to make a title eligible — across every term year. Both are agentic RAG. They feed each other. That closed loop is the product.

Near-miss clusters 💡 Eligibility Reasoner Make-it-eligible levers 📈 Insight Narrator Multi-year visuals agent re-plans · re-retrieves · re-reasons OUTCOMES ✅ ELIGIBLE Sell today 847 ⏳ CONDITIONAL Awaiting date 234 🔮 FORECAST 12-mo pipeline 412 💡 MAKE-IT-ELIGIBLE Near-miss levers 1,089 < 60s full-catalog run · N × M × Y evaluations · re-runs on every catalog change

✅

Eligible

All eligibility conditions pass. Window open right now. Ready to license immediately.

847

titles, per sample deal

⏳

Conditional

Rules pass; the window hasn't opened. Date formulas evaluated — auto-promote when triggered.

234

awaiting trigger dates

🔮

Forecast

Will become eligible once a future release date is met. Pipeline visibility for the next 12 months.

412

12-month horizon

💡

Make-It-Eligible

Near-miss titles with ranked levers — boost theatrical, attach a tier-A cast member, shift a release date, expand territory.

1,089

near-miss titles · ranked playbook

Worked Example · Setting the Stage

From a 94-page contract to one licensed title.

Follow a real Aurora catalog title — FRONTIER QUEST (2022) — through the four rule segments extracted from a CEE Pay-TV output deal. Plain-English story first; the JSON that powers it waits at the bottom.

🎬

FRONTIER QUEST

MPM · F2902700000 · titleId 4

TypeFeature

RatingPG-13

ColourColor

Runtime120 min

OriginU.S.A.

Sales RatingASTRO-MEGA

Module 2 · Live Catalog Record

One title. Ninety-plus structured attributes. Already in the catalog.

When the rule engine runs, every condition in the contract maps to a populated field on FRONTIER QUEST's real title record — theatrical dates, home-video dates, box office, screens, rating, origin, genre. No guessing, no field invention. This is Module 2 doing its job.

US Theatrical2022-02-18

US Home Video2022-05-10

US Box Office$148.6M

WW Box Office$401.7M

US Screens4,275

Budget$100M

GenreAction · Adventure

Earliest ReleaseSouth Korea · 2022-02-09

📜

Source Contract

CEE Pay-TV Output Deal · Territory Group: Baltics & Central Europe

Executed Jan 2022 · Term Year 3 · 94 pages · Four rule segments extracted by Module 1

Segment 1Qualifier

Segment 2Start Date Window

Segment 3Caps

Segment 4Rate Card

How We Made It Eligible

Four Gates. FRONTIER QUEST Clears All Four.

Each gate is a rule category pulled from the contract. We show you the contract clause, then the live evaluation against FRONTIER QUEST's real attributes. Pass all four and the title shows up on the sales dashboard — ready to licence.

Gate 1

Qualifier

Gate 2

Start Date Window

Gate 3

Caps

Gate 4

Rate Card

Gate 1 · Eligibility

Is FRONTIER QUEST the kind of title this deal covers?

What the contract says

"Qualifying Pictures means theatrical motion pictures released theatrically in the United States on or after 1 January 2022, rated PG-13 or milder, and generating reported US box office of at least USD 5,000,000."

Evaluated against FRONTIER QUEST

Is it a Feature?Feature ✓

US theatrical on / after 01 Jan 2022?18 Feb 2022 ✓

Rated PG-13 or milder?PG-13 ✓

US box office ≥ $5,000,000?$148,648,820 ✓

✅ Gate 1 PASSAll four conditions true

Gate 2 · Licence Period

When does the licence window open and close?

What the contract says

"The Pay-TV Licence Period shall commence on the earlier of (a) 100 days after the initial US theatrical release date, or (b) 100 days after the US home-video release date — and shall continue for twelve (12) months."

Evaluated against FRONTIER QUEST

Option A — 18 Feb 2022 + 100 days29 May 2022

Option B — 10 May 2022 + 100 days18 Aug 2022

Earlier of the two29 May 2022 ✓

Window ends (+ 12 months)29 May 2023

✅ Window resolved29 May 2022 → 29 May 2023

Gate 3 · Volume Limits

Is there room for FRONTIER QUEST in Contract Year 3?

What the contract says

"Licensee may exhibit a maximum of twelve (12) Qualifying Pictures per Contract Year, each entitled to up to fourteen (14) linear airings. No more than three (3) ASTRO-MEGA titles may be included in any single Contract Year."

Evaluated against FRONTIER QUEST (Contract Year 3)

Title slot available (of 12)?slot 7 of 12 ✓

14 airings reservable?14 available ✓

ASTRO-MEGA sub-cap (3 max)?2 of 3 used ✓

✅ Within capsslot 7 reserved · 14 airings · ASTRO-MEGA 3/3

Gate 4 · Licence Fee

How much does FRONTIER QUEST licence for?

What the contract says

"Licence fees shall be tiered by US box office: Tier A (≥ $100M) = $2.5M; Tier B ($25M–$100M) = $1M; Tier C (< $25M) = $250K. A territorial multiplier of 1.10× applies to all Baltic territories."

Evaluated against FRONTIER QUEST

US box office $148,648,820Tier A

Tier A base fee$2,500,000

Baltic multiplier× 1.10

Licence fee resolved$2,750,000 ✓

✅ Rate card resolvedTier A · $2.75M

Title	Segmentation	Window Opens	Window Ends	Licence Fee	Airings	Cap Slot
FRONTIER QUEST	Eligible	2022-05-29	2023-05-29	$2,750,000	14	7 / 12

🔬

Deep Dive

The JSON That Made It Happen.

For the technical reader: here's the exact structured output Module 1 extracted from the contract for each of the four rule categories. Each block is the versioned, executable JSON that the rule engine consumes — grounded by the Module 2 reference vocabulary, approved by a human, and auditable end-to-end. Click any card to expand.

Gate 1 · Rule SegmentQualifier — Eligibility Criteria

›

Contract clause (verbatim)

"Qualifying Pictures means theatrical motion pictures (i) produced or acquired by Licensor, (ii) released theatrically in the United States on or after January 1, 2022, (iii) rated PG-13 or milder, and (iv) generating reported US box office of at least USD 5,000,000."

Module 1 extracted JSON

{

"segment": "QUALIFIER",

"condition": "AND",

"rules": [

{ "attr": "titleType", "op": "EQ", "value": "Feature" },

{ "attr": "usTheatDate", "op": "GTE", "value": "2022-01-01" },

{ "attr": "mpaaRating", "op": "IN", "value": ["G","PG","PG-13"] },

{ "attr": "usBoxOffice", "op": "GTE", "value": 5000000 }

]

}

Gate 2 · Rule SegmentStart Date Window — Licence Period

›

Contract clause (verbatim)

"The Pay-TV Licence Period for each Qualifying Picture shall commence on the earlier of: (a) 100 days after the initial US theatrical release date, or (b) 100 days after the US home-video release date — and shall continue for twelve (12) months."

Module 1 extracted JSON

{

"segment": "START_DATE_WINDOW",

"conditionId": 5, // EARLIER_OF

"formulas": [

{ "attr": "usTheatDate", "offset": 100, "uom": "DAY", "territory": "USA(235)" },

{ "attr": "usHvDate", "offset": 100, "uom": "DAY", "territory": "USA(235)" }

"duration": { "value": 12, "uom": "MONTH" }

}

Gate 3 · Rule SegmentCaps — Volume & Exhibition Limits

›

Contract clause (verbatim)

"Licensee may exhibit a maximum of twelve (12) Qualifying Pictures per Contract Year, each picture entitled to up to fourteen (14) linear airings across the Pay-TV window. No more than three (3) ASTRO-MEGA rated titles may be included in any single Contract Year."

Module 1 extracted JSON

{

"segment": "CAPS",

"caps": [

{ "type": "TITLE_COUNT", "max": 12, "perPeriod": "CONTRACT_YEAR" },

{ "type": "AIRINGS", "max": 14, "perTitle": true },

{ "type": "SUB_CAP", "filter": "salesRating == 'ASTRO-MEGA'", "max": 3 }

]

}

Gate 4 · Rule SegmentRate Card — Licence Fee Tiers

›

Contract clause (verbatim)

"Licence fees shall be calculated per Qualifying Picture by tier of reported US box office: Tier A (≥ USD 100,000,000) = USD 2,500,000; Tier B ($25M–$100M) = $1,000,000; Tier C (< $25M) = $250,000. A territorial multiplier of 1.10× applies to all Baltic territories."

Module 1 extracted JSON

{

"segment": "RATE_CARD",

"rateCardType": "TIERED_FLAT_FEE",

"tiers": [

{ "name": "A", "filter": "usBoxOffice >= 100000000", "fee": 2500000 },

{ "name": "B", "filter": "25000000 <= usBoxOffice < 100000000", "fee": 1000000 },

{ "name": "C", "filter": "usBoxOffice < 25000000", "fee": 250000 }

"territoryMultiplier": { "Baltics": 1.10 }

}

Engine 2 · Insights Layer

The dashboard a planner opens every morning.

Engine 2 renders the rule × title × term-year cube as a continuously-updated visual timeline, explains per-title eligibility with the exact reason and term year, integrates with Salesforce, ERP, Tableau and your rights-management system to answer the one question that actually matters: "is this title free to sell — right now, this quarter, next year?"

🗓️

Visual Timeline

Multi–term-year timeline per deal × product category. Title pills colour-coded: Eligible · Conditional · Forecast · Near-miss · Blocked. Amendment markers and cap-usage bars inline. Re-renders on every catalog change.

🔍

Per-Title, Per-Year "Why"

Click any title: get the verdict, the contract clauses that drove it, the evaluated dates, the cap availability, and the rate-card tier — for the exact term year. Every answer traceable to the source clause.

💡

Make-It-Eligible Levers

For every near-miss title: ranked levers (boost theatrical, attach cast, shift date, expand territory), each with its expected revenue impact. Turns CLIP from a report into a revenue-strategy assistant.

🔐

Rights-Availability Check

Integrates with Rightsline, FilmTrack, ERP-RMS, custom systems. Detects "already sold" conflicts before surfacing a title as eligible. Turns rules-eligible into actually-available-to-sell.

☁️

Salesforce · ERP · Tableau

Native connectors: eligible titles auto-create Salesforce Opportunities, rate-card values post to ERP AR, Tableau/Power BI/Looker query the rule × catalog × year cube live. Zero net-new tools for planners.

🔌

API · Webhooks · Export

REST + GraphQL for raw extracted rules and per-title eligibility. Webhooks on eligibility-state change, cap exhaustion, amendment impact. CSV / Excel / PDF export for any slice.

🎯

Engine 2 doesn't just compute eligibility — it shows it, explains it, unlocks it, verifies it can be sold, and delivers it into the systems your sales and finance teams already live in.

That's the difference between a reporting tool and a sales system — and the reason Engine 2 is the insights engine, not just a scoring engine.

Revenue is hiding inside contracts no one can read fast enough.

Two agentic engines. One intelligence loop.

Two Agentic Engines. One Intelligence Loop.

Two Media Companies. One Platform. One Licensed Title.

From a 94-page contract to one licensed title.

One title. Ninety-plus structured attributes. Already in the catalog.

Four Gates. FRONTIER QUEST Clears All Four.

FRONTIER QUEST is ELIGIBLE — licence-ready for the CEE Pay-TV deal.

FRONTIER QUEST surfaces from a 12,000-title catalog scan.

The JSON That Made It Happen.

The dashboard a planner opens every morning.

6-Month Rollout. Quarterly Value Releases.

Three Tiers. One Platform. Grows With You.

The Numbers Investors Need.

Ready to Turn Contracts
into Revenue?

CLIP™

Revenue is hiding inside contracts no one can read fast enough.

Two agentic engines. One intelligence loop.

Two Agentic Engines. One Intelligence Loop.

Two Media Companies. One Platform. One Licensed Title.

From a 94-page contract to one licensed title.

One title. Ninety-plus structured attributes. Already in the catalog.

Four Gates. FRONTIER QUEST Clears All Four.

FRONTIER QUEST is ELIGIBLE — licence-ready for the CEE Pay-TV deal.

FRONTIER QUEST surfaces from a 12,000-title catalog scan.

The JSON That Made It Happen.

The dashboard a planner opens every morning.

6-Month Rollout. Quarterly Value Releases.

Three Tiers. One Platform. Grows With You.

The Numbers Investors Need.

Ready to Turn Contractsinto Revenue?

Ready to Turn Contracts
into Revenue?