Zenith AI Gateway

Zenith AI Gateway — sign in to continue.

SimFictional data
4Demo roles
0Prod systems

Sign in

Accounts are provisioned by an administrator. No self-registration.

Invalid email or password. Try a demo account below.
Email
Password
Verifying credentials…

Prototype accounts — click to fill and sign in

Session expiring

You've been idle. For security this session ends automatically in 15s.

Sign out?

Unsaved filter selections will be cleared. You'll return to the sign-in screen.

Gateway Dashboard
$14M Annual Savings Loading
??
Switch role (demo)
📡 Requests/sec
4,280
Peak: 12,400
🤖 Models Active
8
3 providers
⚡ Cache Hit Rate
34%
$41K saved/mo
💸 Monthly Spend
$1.3M
▼ 40% from $2.5M
🛡 DLP Blocks
142
Today
Provider Distribution (Current)
Azure OpenAI GPT-4o
42%
Fine-tuned LLaMA
38%
Semantic Cache
34%
Embeddings
16%
Live Request Log
[09:14:24] /v1/chat → CACHE HIT → 8ms, saved $0.042
[09:14:23] /v1/completions → llama-3-finserv → 142ms
[09:14:22] /v1/embeddings → ada-002 → 18ms
[09:14:21] /v1/completions → gpt4o fallback (llama timeout) → 420ms
[09:14:20] /v1/chat → DLP scan → clean → llama → 138ms
Model Router Configuration
Active Routing Rules
ConditionRoute ToFallbackCost/1KEnabled
tokens < 500 && task=summarizellama-3-finservgpt4o-mini$0.002
task=complex_analysisazure-gpt4oclaude-3.5$0.015
task=embeddingtext-emb-3-largeada-002$0.00013
semantic_cache_hit=truecache$0.000
compliance_flag=trueazure-gpt4o (audit)none$0.015
Add New Rule
Condition
Route To
Fallback
Cost Savings from Routing
Requests shifted to LLaMA
38% → $18K/mo saved
Cache deflection
34% → $41K/mo saved
Total monthly savings
$59K / month
Semantic Cache Manager
Hit Rate
34%
Target: 40%
Cached Entries
48,200
Active
Tokens Saved
2.8B
This month
Saved Cost
$41K
This month
Query PatternHitsSimilarity ThresholdTTLSaved Tokens
"Summarize Q2 earnings call transcript"2840.9224h142K
"What is our Basel IV capital ratio?"2180.954h109K
"Explain SOFR transition impact"1960.9148h98K
"List high-risk counterparties"1420.971h71K
"Draft regulatory filing boilerplate"1240.9372h62K
DLP / PII Scrubbing Gateway
Requests Scanned
4.2M
This month
PII Blocked
142
Today
Redaction Rate
0.003%
False Positives
0.1%
DLP Policy Rules
Credit Card Numbers (PCI)
Active — BLOCK & LOG
SSN / Tax IDs
Active — REDACT
Account Numbers (ACCT)
Active — REDACT
Employee Names + IDs
Active — REDACT
Insider Trading Keywords
Active — BLOCK & ALERT
App Portfolio — 200+ AI Applications
Total Apps
248
Active
203
Deprecated
45
Consolidated ↓
Savings from Consolidation
$14M
App NameBUModelMonthly CostRequests/dayStatus
FraudDetect ProRiskgpt4o$24,400480,000Active
ComplianceCopilotLegalllama-3-finserv$1,20028,000Active
SupportBot v2Customergpt4o-mini$3,40092,000Active
LegacyAnalyzerITgpt-3.5 (old)$00Deprecated
ShadowReportsUnknownazure-gpt4 (direct)$2,80014,000Shadow
Cost Analysis
Monthly Spend
$1.3M
▼ 48% from $2.5M peak
Annual Savings
$14M
vs unmanaged spend
Apps Deprecated
45
Redundancy eliminated
Business UnitAppsSpendBudgetVarianceTrend
Risk & Fraud48$498K$520K▼ $22K↓ 4%
Customer & CX36$212K$200K▲ $12K↑ 6%
Compliance / Legal28$148K$160K▼ $12K↓ 8%
Research14$187K$180K▲ $7KFlat
Operations / IT22$89K$100K▼ $11K↓ 11%
Shadow AI4$166K$0UnauthorizedEscalated
Snowflake Immutable Audit Trail
TimestampEventAppUserModel UsedDLP ActionHash (Snowflake)
09:14:24REQUESTFraudDetect Prosys-agentazure-gpt4oCleana8f3b2c1d4e5
09:14:22DLP_BLOCKShadowReportsr.chenazure-gpt4 (direct)SSN blockedc2e9d4f8a1b3
09:14:18REQUESTComplianceCopilotl.torresllama-3-finservCleanf7a1c8b2d3e4
09:14:15CACHE_HITSupportBot v2sys-agentcacheN/A3b8e2f1a7c9d
Compliance Posture
FrameworkControlsPassedEvidence ItemsStatus
HIPAA — PHI Protection141442 itemsCompliant
SOC 2 Type II — AI Systems181886 itemsCompliant
FINRA — Supervisory Controls10931 items1 Gap
OCC AI Risk Guidance8824 itemsCompliant
GDPR — Data Processing121236 itemsCompliant
Value Drivers — ROI Attribution
💰 Total Annual ROI
$14M
12-month payback
🔄 Cost Reduction
~40%
$2.5M → $1.3M/mo
🗑 Apps Deprecated
45
$5.4M saved/yr
⚡ Cache Savings
$492K
Annual
ROI by Initiative
App consolidation (45 deprecated)
$5.4M/yr
Model routing (LLaMA vs GPT-4)
$4.2M/yr
Shadow AI elimination
$2.4M/yr
Semantic cache savings
$1.5M/yr
Compliance automation
$0.5M/yr
Shadow AI Monitor
4 Unapproved Apps Detected
App NameDepartmentAPI UsedEst. Monthly CostRiskDetectedAction
ShadowReportsFinance — R. ChenAzure OpenAI (direct)$2,800HighJun 20
QuickSummarizeLegal — unknownChatGPT API$420MediumJun 18
TradingBotResearch — T. MorelAnthropic API$1,200HighJun 17
MeetingAIHR — UnknownOpenAI Whisper$180LowJun 15