AllCaps Under the Hood
Most procurement teams face a brutal reality: hundreds of contracts, tiny teams, zero bandwidth for negotiation. Here's how we built an AI system that doesn't just assist—it actually negotiates.
Inside AllCaps: How 4 Agents Work Together to Capture Contract Value
Most procurement teams face a hard reality: hundreds of contracts, small teams, and limited bandwidth to operationalize contract terms. The result isn’t just missed renewals—it’s unclaimed value embedded in everyday contracts.
After building AllCaps and supporting real contract engagements, one thing became clear: the most effective systems aren’t built around a single agent. They’re built as specialized teams that identify opportunities, assess leverage, and execute the right capture path.
That’s the model we’ve built at AllCaps.
Meet the Team: 4 Agents That Work Together to Realize Contract Value
🔍 Scout: Contract Intelligence
What Scout Does: Ingests and analyzes contracts across the organization to surface value opportunities, risks, and trigger events.
Core Capabilities:
- Contract Parsing: Extracts key terms, pricing structures, obligations, and rights from PDFs/DOCX
- Trigger Detection: Identifies renewal windows, usage thresholds, SLA breaches, and other value moments
- Risk & Fee Identification: Flags hidden fees, unfavorable clauses, and exposure
- Baseline Establishment: Captures current pricing, terms, and entitlements
Why This Matters: Contract value is scattered across systems and teams. Scout creates a single, normalized view so opportunities don’t go unnoticed.
🧠 Strategist: Value & Leverage Assessment
What Strategist Does: Evaluates each opportunity to determine whether value can be captured through enforcement or requires vendor engagement.
Key Dimensions Assessed:
- Enforceability: Are there contractual rights that can be exercised as-is?
- Leverage: Switching costs, alternatives, timing, and buyer impact
- Market Context: Benchmarking, vendor position, and competitive dynamics
- Strategic Importance: Account value to the vendor and relationship context
- Evidence Strength: Data required to support enforcement or negotiation
The Output: A clear execution recommendation:
- Enforce existing terms
- Engage vendor for concessions or changes
- Defer or monitor
Example: In a SaaS agreement, Strategist identified enforceable SLA credits and pricing protections, reducing spend without renegotiating core terms.
📋 Playbook: Execution Planning
What Playbook Does: Converts strategy into execution-ready plans teams can act on immediately.
What Playbooks Include:
- Value Cards: Rights, fees, or concessions with evidence and expected impact
- Execution Path: Enforce vs. engage recommendation
- Language & Templates: Emails, call scripts, and escalation paths
- Calculators: Quantify impact of credits, pricing changes, and concessions
- Fallbacks: Alternative paths if resistance is encountered
Why This Works: Teams get clear, practical guidance without weeks of analysis.
🤝 Calm™: Coordination & Decision Support
What Calm Does: Supports execution by managing approvals, adapting strategy, and coordinating stakeholders.
Key Functions:
- Presents recommendations and explains tradeoffs
- Routes decisions to Finance, Legal, IT, and leadership
- Analyzes vendor responses and updates next steps
- Tracks progress and documents outcomes
Why Calm Matters: Execution is dynamic. Calm ensures strategies evolve as conditions change and decisions stay aligned.
How the Agents Work Together
Stage 1: Contract Ingestion (Scout)
- Normalize contracts and extract rights, risks, and triggers
Stage 2: Opportunity Assessment (Strategist)
- Determine enforceability, leverage, and priority
Stage 3: Execution Planning (Playbook)
- Generate actionable playbooks and evidence packs
Stage 4: Execution & Adaptation (Calm)
- Enforce terms or engage vendors
- Adapt based on responses and new information
Stage 5: Validation (Scout + Calm)
- Document realized value and update baselines
Why This Architecture Works
-
Rights-First Execution Value is captured by exercising entitlements first, not defaulting to negotiation.
-
Specialization Over Generalization Each agent focuses on what it does best—analysis, assessment, planning, or coordination.
-
Human Control, AI Support Teams approve strategies and outcomes. The system shows its work.
-
Repeatable, Scalable Results The same rigor applies across large and small contracts.
What This Means for Procurement Teams
- Fewer missed entitlements and credits
- Less time spent manually reviewing contracts
- Clear prioritization of where to act
- Measurable contract value captured throughout the year
Closing
AllCaps is designed to help teams move from knowing what’s in their contracts to actually realizing the value inside them.
If you’d like to see how these agents would approach a real contract, I’m happy to walk through an example.