Parinita AI Edge Agent Portfolio
21+ Sovereign Agents. Every One Cryptographically Identified and On-Chain Governed.
Platform agents. Vertical agents. Infrastructure ops agents. Every action Chrysalis-anchored.
The AI-intelligence layer of the Parinita platform expressed as 21+ named agents, organized into three categories: Platform Agents embedded in named product lines, Vertical Agents for industry-specific intelligence, and Infrastructure Operations Agents for autonomous infrastructure management.
Enterprise AI Agents Have No Cryptographic Identity and No Runtime Governance
AI agents deployed in enterprise environments operate with no cryptographic identity, no runtime enforcement of permissions, and no tamper-proof audit trail of their actions.
When an AI agent takes an action — accessing a file, calling an API, executing a workflow — there is typically no cryptographic proof of which agent did it, under what permissions, and whether those permissions were valid at the time.
-
Agents Have No Cryptographic Identity
When an AI agent acts on a system, there is typically no cryptographic proof of which agent did it, under what warrant, and whether the warrant was valid at the time of action.
-
Agent Permissions Are Not Enforced at Runtime
Documented agent permissions exist in policy documents. Without hardware-level BPF map enforcement, an agent can reach any resource its network path allows.
-
Vertical AI Has No Regulated Data Handling
Industry-specific AI agents from cloud providers operate outside regulated data boundaries — PHI, ITAR-controlled data, and PCI data cannot travel to their inference endpoints.
Three agent categories. Every action on-chain.
Every Parinita agent operates under a Sentry-issued Ed25519 identity, a Policy Warrant enforced at the NIC, and a Chrysalis audit trail. This is where sovereign infrastructure becomes sovereign intelligence.
The Sentry-Overwatch-Chrysalis triangle governs the entire portfolio simultaneously — artifact governance at publication, runtime enforcement at the NIC, behavioral monitoring continuously.
-
Platform Agents (8)
Accord (Nexus governance), Sage (Pulse analytics), Q (CRM), Warden (Chrysalis policy), Investigator (provenance), Dispatch (Conduit routing), Atma (digital twin), Vaak (sovereign voice TTS).
-
Vertical Agents (8)
Counsel (Legal, $175/seat), Guardian (Healthcare, $350/seat), Ledger (Financial Services, $500/seat), Forge (Manufacturing, $250/seat), Shield (Cybersecurity, $400/seat), Lens (Media, $200/seat), Atlas (Government, $800/seat), Campus (Education, $50/seat).
-
Enclave Agents (2)
Stratum — sovereign object storage engine in Rust. Flash — Enclave orchestration agent handling the write pipeline, replication, and ransomware threat correlation.
-
Infrastructure Ops Agents (10+)
Pulse Ops 1-3 for telemetry and auto-remediation, plus dedicated agents for capacity, network, security, provisioning, cost, and patching across all 101 POPs.
-
Hardware-Enforced Identity
Every agent runs under a Sentry-issued Ed25519 identity. eBPF/XDP on ConnectX-7 NICs drops unauthorized packets before the kernel ever sees them.
-
On-Chain Governance
Every agent operates under a Policy Warrant compiled into BPF maps. Warrant changes anchor on Chrysalis. Behavioral baseline anomalies route to Overwatch.
-
Chrysalis-Anchored Audit Trail
Every agent action that matters anchors on Chrysalis — tamper-proof by architecture. Not a log. A blockchain-verified record of what happened, when, under what warrant.
-
Sentry-Overwatch-Chrysalis Triangle
Every agent in the portfolio passes through artifact governance (Sentry), runtime behavioral monitoring (Overwatch), and tamper-proof audit (Chrysalis) simultaneously.
-
Conduit Inference Routing
Every agent's inference requests route through Conduit — the sovereign multi-model inference router — ensuring the right model, the right plane, and the right jurisdiction for every request.
The Plane Model
Orchestra introduces "planes" — logical groupings of hardware optimized for a specific workload class. Unlike Kubernetes node pools, planes represent fundamentally different hardware architectures with different drivers, network requirements, and scheduling semantics.
The plane model is what makes Orchestra different from every other orchestration tool. Kubernetes sees nodes. Orchestra sees purpose-built hardware tiers and routes workloads accordingly.
Identity, warrant, behavior. Every agent, every action.
Every agent is grounded in Crucible per-packet identity, Corridor cross-POP communication, Chrysalis on-chain governance, and Conduit inference routing. Every action passes through the Sentry-Overwatch-Chrysalis triangle.
- 01Agent identitySentry-issued Ed25519 keypair per agent — cryptographic identity at the NIC, not asserted by the agent itself. Cannot be spoofed.
- 02Policy WarrantPer-agent permissions compiled into BPF maps distributed to ConnectX-7 NICs. The enforced policy — not the documented one.
- 03Artifact governanceFive-track Sentry scan pipeline — static code, gVisor behavioral sandbox, publisher reputation, vector similarity, version delta — before any agent can be published or deployed.
- 04Inference routingEvery inference request routes through Conduit to the optimal model, plane, and jurisdiction. No agent makes direct model API calls outside the sovereign fabric.
- 05Behavioral monitoringSentry continuously baselines tool calls, plane access, data volumes, and inter-agent invocations. Anomalies route to Overwatch for cross-domain response.
- 06Vertical pricingVertical agents price per seat per month: Campus $50, Counsel $175, Lens $200, Forge $250, Guardian $350, Shield $400, Ledger $500, Atlas $800.
- 07Jurisdiction handlingVertical agents that process regulated data (PHI, ITAR, PCI) run at the jurisdiction POP. Sovereign AI Processing keeps regulated data in bounds.
- 08Ops agent autonomyInfrastructure ops agents (Pulse Ops, capacity, network, security, provisioning) operate autonomously across 909+ clusters — no human approval required for routine operations.
import agents
client = agents.Client(seat_token='your_seat_token')
# List active agents and their warrants
for agent in client.agents.list():
print(f"Agent: {agent.name}, Category: {agent.category}, Warrant: {agent.warrant_hash}")
# Run a vertical agent task
result = client.agents.run(
agent='counsel',
task='analyze_contract',
document_id='contract_2026_Q2',
jurisdiction='us-east'
)
print(f"Finding: {result.summary}, Chain: {result.chrysalis_tx}")Proven at scale. Not in a lab.
Parinita AI Edge is the production deployment of the Parinita platform and the largest heterogeneous AI infrastructure deployment in the United States.
Parinita AI Edge
The most complex heterogeneous AI infrastructure in the United States. 101 sites, 9 planes, 12,000+ nodes, 4 accelerator vendors, dual network fabrics, four-layer tenant isolation — all through a single sovereign control plane.
Network & Security Infrastructure
- Multi-vendor acceleratorsFour accelerator vendors — Intel Habana, NVIDIA, AMD, Qualcomm — orchestrated through one control plane with unified scheduling, monitoring, and lifecycle management.
- Dual-fabric networkingCisco production fabric and Arista GPU backend fabric operating as a coordinated system, bridged by identity-aware routing.
- Nationwide scale101 sites across 42 U.S. states, each operating autonomously with a local control agent and a sovereign cross-site routing plane.
- Multi-tenant isolationFour-layer defense-in-depth: VXLAN VNIs, identity-routing, Palo Alto firewalls, and Cilium eBPF — validated across every plane and site.
- Compliance readinessFIPS 140-2 at launch, with FedRAMP Moderate, CJIS, and IL4/IL5 certification paths active through Parinita compliance profiles.
- Sub-millisecond routingEvery request classified and dispatched in under 1ms, enabling real-time SLA enforcement without perceptible overhead.
Sovereign intelligence for every industry and every infrastructure function.
Vertical agents bring domain expertise to regulated workflows — Counsel for legal, Guardian for healthcare, Ledger for financial services — all with the same cryptographic identity and governance as every other Parinita agent.
Infrastructure ops agents handle capacity, network, security, provisioning, cost, and patching autonomously — reducing the operational burden of managing 909+ clusters across 101 POPs.
-
Legal Workflows
Counsel agent — contract analysis, M&A due diligence, regulatory mapping. $175/seat. Every finding anchored on Chrysalis for court-admissible chain of custody.
-
Healthcare Operations
Guardian agent — clinical decision support, PHI-safe summarization, HIPAA-compliant patient communications. $350/seat. PHI stays in jurisdiction by architecture.
-
Financial Services
Ledger agent — fraud detection, KYC/AML, regulatory filings. $500/seat. Every financial AI action Chrysalis-anchored for SEC and FINRA audit trails.
-
Infrastructure Automation
Ops agents handle capacity, network, security, provisioning, cost, and patching autonomously across 909+ clusters — reducing the operational burden of multi-silicon AI edge management.
-
Government and Defense
Atlas agent — citizen services automation, ITAR-compliant classified document handling, geospatial intelligence. $800/seat. Operates within ITAR-compliant entity boundaries.
Deployment Models
Platform agents activate automatically with their parent product. Vertical agents provision at the seat level — per-agent pricing applies.
Platform agents (Accord, Sage, Q, Warden, Investigator, Dispatch, Atma, Vaak) activate automatically when their parent product is provisioned. No separate setup.
Vertical agents provision at the seat level. Pricing from $50/seat (Campus) to $800/seat (Atlas). Multiple vertical agents can run per seat.
Pulse Ops 1-3 and all infrastructure ops agents activate as part of the platform fabric. No per-agent provisioning — they monitor and remediate automatically.
Talk to Us
Our AI team can walk through the agent portfolio and vertical agent pricing for your specific industry and workflow requirements.