Live
Home
Who It's For
The Problem
Theia Assist
How It Works
FAQ
Get Started
Early Access Pilot operators onboarding

The accountability layer for dark fiber

The source of truth
for dark fiber.

Critical event records, incident lifecycle, and continuous monitoring in one carrier-grade system of record. Operators issue the proof. Enterprises see what they're paying for. One platform. One record.

Operator-issued. Customer-readable. Independently verifiable.

Multi-tenant from day one  ·  Carrier-grade  ·  No hardware lock-in

Theia Fiber Dashboard: real-time fiber monitoring and SLA governance

Two sides of the same record.
Pick yours.

For network operators

Prove it before you have to defend it.

Customer disputes, audits, SLA credit reviews. Every escalation in dark fiber operations comes back to one question: what actually happened, and can you prove it? Without a system of record, the answer is reconstruction: engineering hours stitching together exports, optical traces, and timestamps after the fact. Theia replaces that with a signed artifact. Every optical event becomes an attributed, time-stamped record automatically, in a chain that holds up in front of customers, regulators, and finance.

Single source of truth
Audit-ready by default.

Every optical event becomes a signed, attributed CER. No reconstruction. Ready for customers, auditors, and finance the moment it's issued.

Tenant scoping
Each tenant sees their own plant.

Customer-facing views scoped per circuit. They get clarity on the fiber they pay for. You stay in control of what gets shared.

Carrier scale
Built for portfolios.

Multi-tenant dark fiber by design. Hundreds of circuits across thousands of spans, on one carrier-grade record.

Request a Pilot →
For enterprise customers

See your leased plant in real time.

You signed a contract for fiber you can't see. Theia gives you a continuous read on every circuit you lease, with full visibility into what's happening and where.

Always-on visibility
Read every leased circuit, live.

Continuous monitoring on the plant you pay for. No carrier handover, no filed ticket required to see what's actually happening.

Shared record
Same data, both sides.

The signed record your carrier references is the one you see. Diagnose together, not in parallel from different systems.

Audit-ready reporting
Snapshot-grade evidence.

Built for finance, procurement, and risk. Every SLA conversation backed by data both sides already agree on.

Request a Pilot →

You know the circuit is down.
You don't know why, where,
or for how long.

One scenario. Two realities. A fiber break at 3 AM. Theia picks it up at the optical layer. Detection is the easy part. The harder part comes after. Without a governance layer, what follows is hours of reconstruction, fragmented evidence, and a downtime number that depends on whose log you trust. With Theia, the break becomes a signed, attributed, reconciled record automatically.

Incident scenario · Complete fiber break · SP-14 · 03:14 UTC · Circuit NYC-DCA-042 ⚠ Complete outage
Without a governance layer
hours of reconstruction · incomplete record
hrs / days
03:14:09 UTC
Router alarm fires. Cause unknown.
The circuit is down. Whatever data exists lives in separate systems. Network logs in one place, optical telemetry in another, ticketing somewhere else. Nothing is correlated. Nothing is attributed to a contract or counterparty.
Uncorrelated
03:14 to 05:30 UTC
Internal troubleshooting begins. Hours pass.
Engineers work through their own gear first. Eventually the problem is declared external. But there's no commercial context attached. Nothing tying the event to a specific span, MSA, or SLA commitment. It's still a complaint, not a contractual event.
No attribution
~05:30 UTC
Trouble ticket opened. The other side asks for proof.
The counterparty wants specifics: when did it break, where, what happened, how long was it down. You have fragments. They have their own fragments. Neither version is authoritative. Neither is signed.
No evidence package
~06:30 UTC
Truck rolls. Evidence is still being assembled.
Field crew heads out. Meanwhile, someone on your team is still piecing together the incident timeline from syslogs, emails, and call records. That work continues long after the fiber is repaired.
Manual reconstruction
Fiber restored
Circuit's back. Now the argument starts.
How long was it down? Who's accountable? What does the contract say? Both sides produce different numbers from different systems. No signed record means responsibility is negotiated, not calculated. SLA credits get argued down. The relationship takes the hit.
No signed record · Credits negotiated
With Theia
signed record · attributed · reconciled
min not days
03:14:09 UTC
Theia picks up the break and logs it as a contractual event, not just an alarm.
The moment the event fires, Theia correlates it against your commercial data model. The affected span maps to its circuit, which carries the MSA, availability commitment, and credit terms. The incident isn't "something broke." It's a contractual event with a known, accountable party. Whoever the data points to.
Correlated · Attributed
03:15:00 UTC
Evidence package assembled automatically.
Break timestamp, location, baseline delta, and contractual context are bundled the moment the event fires. Nothing to gather, nothing to reconstruct. Send it when you're ready: a complete package, not a complaint.
Evidence ready
03:15:02 UTC
Structured incident opens. SLA breach clock starts.
Theia opens an immutable incident record. The SLA clock runs against the actual commercial terms. Not a generic threshold. The exact contract. Every event from this point is timestamped and logged into the record automatically.
Auto · Breach clock live
03:16:00 UTC
The counterparty sees the same record you do.
Dual-principal reconciliation: both sides read from the same authoritative source. No parallel investigations, no conflicting timelines, no "our logs say different." One record, shared, timestamped, signed.
Shared source of truth
Fiber restored
Critical Events Record issued. Incident closed.
On restoration, Theia produces a Critical Events Record: cryptographically signed, contractually attributed, reconciled across both parties. Actual downtime measured against actual SLA terms. Credit exposure calculated, not argued. Audit-ready the moment it's issued.
✓ Signed · Reconciled · Audit-ready
3h+
Troubleshooting the wrong layer
<2 min
Theia confirms root cause
0
Manual steps to generate evidence
70%
Reduction in MTTR
95%
Faster root cause identification
±2m
Fault location precision
No ambiguity. No disputes. Responsibility is clear.
Early Access

Not a chatbot.
A fiber network teammate.

Most AI tools answer one question at a time. Theia Assist handles the whole investigation: pulling span history, correlating alarm sequences, checking SLA thresholds, and returning a finding with supporting evidence tied to the underlying data. You ask. It works.

Autonomous investigation
Formulates a multi-step plan, executes queries, and returns findings without you directing each step.
Fiber-domain expertise built in
Trained on optical telemetry workflows including OTDR baselines, insertion loss deltas, and span degradation patterns, not generic IT ops.
Evidence-linked answers
Every response cites the underlying telemetry and timestamps so findings are traceable, not just plausible.
SLA-aware from the start
Every finding is evaluated against your active SLA thresholds, so Theia tells you not just what happened, but whether it counts.
Theia AI
Online
Today

Span serving Tenant B between Newark and Manhattan has deviated from baseline. Currently within SLA threshold. Pattern flagged for investigation before customer impact materializes.

Tenant B: Watch EARLY WARNING
CustomerTenant B
CircuitNewark → Manhattan
Signal trendDeviated from baseline
Current statusWithin SLA, no breach
Est. time to impactMonitoring

Signal to evidence.
Automatically.

Theia sits between your fiber infrastructure and your SLA obligations, converting every optical event into a verified, enforceable record. A single source of truth, automatically generated.

Your Fiber Infrastructure
Optical signals measured continuously
SLA governance and real-time optical monitoring, 24/7.
Theia Platform
AI correlates, escalates, and documents
Root cause confirmed, incident opened, SLA clock running.
Critical Events Record (CER)
Signed, timestamped, delivered automatically
Every incident produces one, no exceptions.
Dispute resolved
Case closed before the argument starts
Whoever needs the evidence has it, no negotiation.
<3 mo
Typical payback period
100%
Of SLA events produce a Critical Events Record
0
Manual steps to generate a CER
FAQ

Questions, answered.

Twenty-six questions on platform mechanics, telemetry, the Critical Events Record, SLA math, security posture, Theia Assist, and deployment. Click any question to expand.

Platform
What is Theia Fiber?

Theia Fiber is a SaaS monitoring and governance platform for dark fiber operators, carriers, and data center providers. The platform binds optical telemetry from Adtran inline OTDR hardware to the SLAs and contracts that govern fiber leases, producing an evidentiary record both the operator and the tenant customer can read from. Core capabilities include continuous span monitoring, SLA breach calculation against actual commercial terms, and the Critical Events Record (CER) for incident audit and dispute resolution.

Who uses Theia Fiber?

Dark fiber operators, carriers, and data center providers on one side, and their tenant customers on the other. Both parties have logins and read the same telemetry record from the same platform.

What is the dual-principal trust model?

The provider and the tenant both view the same record and information: the same span loss baseline, the same alarm timeline, the same CER. Neither side controls the record the other side reads. That separation is the trust mechanism.

How is Theia different from an NMS?

An NMS detects faults. Theia attributes them. The output is contractual evidence: which span failed, which circuit was affected, which SLA was breached, what credit is owed. Detection is table stakes; attribution is the product. Theia coexists with your NMS rather than replacing it.

Telemetry & CER
How does Theia see anything on dark fiber?

Dark fiber does not self-report. Adtran's inline OTDR injects a probe wavelength on the strand and measures the returned signal: insertion loss, reflectance, baseline drift, fault distance. The probe runs on a separate wavelength from any tenant traffic, so monitoring continues whether the strand is dark or actively carrying customer service. Theia consumes that telemetry continuously, binds it to circuits and contracts, and produces the audit record. The hardware produces the photons; Theia produces the accountability.

What is a Critical Events Record?

A signed, immutable record of a fiber event that materially affects a circuit. It contains the alarm chain, affected spans and circuits, baseline versus measured loss, the SLA terms in force at the time, the breach calculation, and the resulting credit if applicable.

Is the telemetry record immutable?

Yes. Alarm states are written only by the Beacon. The UI cannot mutate alarm or CER records, and span geometry is locked after first save. Disputes happen through structured workflow, never by editing history.

What hardware does Theia support?

Adtran inline OTDR today. The hardware injects the probe wavelength on the strand and produces the optical telemetry Theia consumes. Theia does not generate optical signal itself. Vendor expansion is on the roadmap; the platform is hardware-agnostic by design, so adding new vendors does not require rearchitecting the data model or the CER format.

Will Theia support other monitoring vendors?

Yes. Adtran inline OTDR is our launch integration and additional vendor integrations are planned as the ecosystem expands. The Beacon architecture and CER data model are vendor-agnostic by design, so new vendors plug into the existing pipeline without changes to circuit binding, SLA math, or the audit record. Customers running mixed plant should expect coverage to grow over time.

Does Theia store raw OTDR traces?

Yes. Traces are retained in their native SOR-compatible format and viewable in the Optical Analysis console. Baselines are captured at provisioning and used as the reference point for all deviation math.

SLA
How are SLA breaches calculated?

Each circuit is bound to its commercial SLA terms: uptime targets, credit schedule, dedup window, grace policy, and chargeable states. When an alarm chain affects a circuit, accumulation math runs against those terms. Every step is auditable.

Can Theia work with our existing contracts?

Yes. SLA terms are configured per circuit, not from a platform default. Existing master service agreements, custom credit schedules, and grace policies are modeled in the SLA Contracts console.

What happens when a breach is recorded?

A CER is generated and made available to the provider. The provider can share it with the tenant at any time. Once shared, both parties view the same record and the same credit calculation. Whether the credit auto-applies or routes for review is a customer-side policy choice, not a platform default.

Can a tenant dispute a CER?

The CER is the evidence. Disputes are about interpretation, not data. Both sides see the same alarm timeline and the same SLA terms, which is what surfaces the disagreement clearly enough to resolve.

Security
Where is customer data stored?

Two options. SaaS is cloud-hosted with strict tenant isolation. On-premise keeps the data inside the customer environment under the customer's data controls. No tenant ever queries another tenant's records in either model.

Do we need to open inbound firewall ports?

No. The Beacon initiates an outbound TLS session to Theia. There is no inbound listener on customer infrastructure and no inbound NAT rules to manage. The connection model is outbound-only, mutually authenticated, and follows standard zero-inbound deployment patterns.

What about BEAD compliance?

Theia produces evidence aligned with BEAD reporting requirements. Span geometry, monitoring continuity, and outage records are exportable in the formats program officers expect to see.

Theia Assist
What can Theia Assist do?

Three things at launch. Answer plant queries in plain English. Alert proactively when measurements drift from baseline. Narrate alarms with diagnostic context. It is read-only and reasoning-only.

Can Theia Assist run scans or restart equipment?

No, by design. Theia Assist will not initiate active operations on customer infrastructure, ever. This is a deliberate trust and liability decision, not a roadmap deferral.

Does Theia Assist see our data?

It reads context relevant to the question being asked, scoped by the user's tenant and role. Conversations are not used for model training and customer data is not shared across tenants.

Deployment
What is the Beacon and where does it run?

The Beacon is the on-site agent that talks to your Adtran inline OTDR. It reads telemetry, normalizes it, and forwards it to Theia over an outbound session it initiates from inside your network. It runs on a small VM or container in your environment and one Beacon typically covers a site or region.

Can Theia be deployed on-premise?

Yes, for compliance, data sovereignty, or regulatory reasons. The same Theia stack runs in an isolated environment under the customer's control with the same data model, the same isolation guarantees, and the same CER format. The difference is operational: software ships as signed releases on a customer-controlled cadence rather than continuous SaaS rollout.

How long does onboarding take?

Span and port import from existing records typically completes within hours. Beacons download and deploy in under 30 minutes per site as a small VM or container. Baseline capture happens as the Adtran inline OTDR connection comes online.

Does Theia replace my existing fiber monitoring tools?

No, Theia ingests data from the OTDR scanners, ALM probes, and chassis already in operation. The platform adds a governance and accountability layer above those tools. If a fiber operator already runs continuous monitoring on its plant, Theia integrates with that telemetry stream rather than duplicating it. There are no required hardware swaps or rip-and-replace projects.

How do upgrades work?

For SaaS, features ship continuously with no customer-side install, no version freeze, and no upgrade window to coordinate. For on-premise, signed releases ship on a customer-controlled cadence so you decide when to upgrade.

Can we run a pilot?

Yes. Pilots scope to a defined plant footprint with a small number of tenant customers and run for an agreed period before commercial conversion.

One record. Both sides.
No reconstruction required.

Operators issue the proof. Enterprises see the truth. Both sides work from the same signed record. Join the operators and enterprises piloting Theia today.

No spam. Pilot access only. Unsubscribe anytime.

Carrier-grade
30-minute deployment
No hardware lock-in