Alika Aurora • Ops Control Center

Secure sign in

Protected by multi-factor authentication

OPS CONSOLE

Central hub for controlling, validating, certifying and promoting system changes.

OpenAI Balance

View Balance ↗

Authoritative balance at OpenAI

Backend Health Monitor

Read-only VPS telemetry · refreshes every 15 seconds

CPU—

Waiting for telemetry

Memory—

Waiting for telemetry

Swap—

Waiting for telemetry

Disk—

Waiting for telemetry

Capacity Summary

Current safe operating picture

Users online—

AI slots processing—

Engineering queue—

Host load / core—

Model memory—

AI failures · 15 min—

NATE trading heartbeat—

App trading heartbeat—

Active App users—

Real App users—

Test users online—

Waiting accounts—

Longest scan delay—

Average cycle time—

Active strategies—

Broker failures—

Certified App ceiling—

Registered App users—

Registration places—

Next load test—

OpenAI Fallback Monitor

BossAI and Ella only · Nate remains permanently local

Live OpenAI requests—

Waiting for fallback telemetry

Users today—

Tokens today—

Input / output—

Last activity—

View activity opens the read-only BossAI chats and engineering runs that used OpenAI.

BossAI Runtime

Local model health, workload and response activity

Waiting for model telemetry —

Model service—

Process CPU—

Process memory—

Context ceiling—

Model slots—

Engineering runs—

Responses · 15 min—

Failures · 15 min—

Users Online

Real authenticated presence plus enabled App load testers

— active users

User	Access	Current area	Last activity
Waiting for presence telemetry

OPS/L5 only · credentials, tokens, MFA codes and IP addresses are never displayed.

Stage	Local AI model	VPS / compute	Memory / disk	AI capacity	Planning population
CurrentFine-tuning	Qwen2.5-Coder 7B InstructQ4_K_M · current model	16 vCPUCPU-only inference	16 GB RAM300 GB NVMe	1 slot8K context	APP: up to 10NATE: up to 10
LaunchSmall production	Qwen2.5-Coder 7B InstructQ4_K_M	24 vCPU1 × 24 GB GPU	64 GB RAM500 GB NVMe	4 slots8K–16K context	APP: ~1,000NATE: ~100
GrowthImproved reasoning	Qwen2.5-Coder 14B InstructQ4_K_M	32 vCPU1 × 48 GB or 2 × 24 GB GPU	128 GB RAM1 TB NVMe	8 slots16K–32K context	APP: ~10,000NATE: ~500
RegionalHigh capability	Qwen2.5-Coder 32B InstructQ4_K_M	48–64 vCPU1 × 80 GB or 2 × 48 GB GPU	256 GB RAM2 TB NVMe	16 slots32K context	APP: ~50,000NATE: ~2,000
TargetAlika Aurora scale	Qwen2.5-Coder 32B serving clusterReplicated Q4_K_M workers	64+ vCPU total2+ × 80 GB GPU across inference nodes	256–512 GB RAM4 TB NVMe + backups	32+ slotsQueue + autoscaling required	APP: ~100,000NATE: ~5,000