MEDIUM
ai_ops_weekly_digest
7× FIRED
3d ago
AI Ops Weekly Digest — Week of April 27, 2026
============================================================
COST STATUS: ON TRACK
MTD: $278.78 | Projected: $309.76 | Budget: $1,500.00 (21%)
7-day avg: $10.37/day
MODEL EFFICIENCY:
Opus ratio: 100% | Wasteful sessions: 10 | Est. waste: $70.12
COST ALLOCATION:
RGDM (internal): $2,125.91 (26 sessions)
SESSION QUALITY: HEALTHY
Avg score: 91/100 | Cache hit: 79.1% | Flagged: 0/26
OVERHEAD TREND: STABLE
Latest avg input: 172,046 tokens/session | WoW: -91.1% | Consecutive growth: 2 weeks
RECOMMENDATIONS:
1. [HIGH] Route trivial sessions (<5 min) to Sonnet/Haiku instead of Opus (est. $280.48/mo savings)
2. [MEDIUM] Investigate low cache hit rate — may indicate context management issues
ALERTS:
[MEDIUM] (efficiency) Opus ratio 100% exceeds 25% threshold
[MEDIUM] (efficiency) Estimated $70.12 waste from 10 trivial Opus sessions
[LOW] (allocator) RGDM (internal) accounts for 100% of AI spend — consider cost optimization
OUTCOME TRACKING:
[=] Opus Ratio: 100% (was 100%, +0.0% — unchanged)
[-] Est. Waste: $70.12 (was $0.58, +11989.7% — regressed)
[-] Budget Pacing: 21% (was 21%, +0.5% — regressed)
[+] Avg Quality: 91/100 (was 89/100, +1.6% — improved)
[+] Cache Hit: 79.1% (was 76.9%, +2.9% — improved)
[+] Overhead: 172,046 tok (was 1,098,002 tok, -84.3% — improved)
4-week trends:
Budget Pacing: 71% → 17% → 19% → 21% (improving)
Opus Ratio: 100% → 100% → 100% → 100% (worsening)
▾
AI Ops Weekly Digest — Week of April 27, 2026 ============================================================ COST STATUS: ON TRACK MTD: $278.78 | Projected: $309.76 | Budget: $1,500.00 (21%) 7-day avg: $10.37/day MODEL EFFICIENCY: Opus ratio: 100% | Wasteful sessions: 10 | Est. waste: $70.12 COST ALLOCATION: RGDM (internal): $2,125.91 (26 sessions) SESSION QUALITY: HEALTHY Avg score: 91/100 | Cache hit: 79.1% | Flagged: 0/26 OVERHEAD TREND: STABLE Latest avg input: 172,046 tokens/session | WoW: -91.1% | Consecutive growth: 2 weeks RECOMMENDATIONS: 1. [HIGH] Route trivial sessions (<5 min) to Sonnet/Haiku instead of Opus (est. $280.48/mo savings) 2. [MEDIUM] Investigate low cache hit rate — may indicate context management issues ALERTS: [MEDIUM] (efficiency) Opus ratio 100% exceeds 25% threshold [MEDIUM] (efficiency) Estimated $70.12 waste from 10 trivial Opus sessions [LOW] (allocator) RGDM (internal) accounts for 100% of AI spend — consider cost optimization OUTCOME TRACKING: [=] Opus Ratio: 100% (was 100%, +0.0% — unchanged) [-] Est. Waste: $70.12 (was $0.58, +11989.7% — regressed) [-] Budget Pacing: 21% (was 21%, +0.5% — regressed) [+] Avg Quality: 91/100 (was 89/100, +1.6% — improved) [+] Cache Hit: 79.1% (was 76.9%, +2.9% — improved) [+] Overhead: 172,046 tok (was 1,098,002 tok, -84.3% — improved) 4-week trends: Budget Pacing: 71% → 17% → 19% → 21% (improving) Opus Ratio: 100% → 100% → 100% → 100% (worsening)
AI Ops Weekly Digest — Week of April 27, 2026 ============================================================ COST STATUS: ON TRACK MTD: $278.78 | Projected: $309.76 | Budget: $1,500.00 (21%) 7-day avg: $10.37/day MODEL EFFICIENCY: Opus ratio: 100% | Wasteful sessions: 10 | Est. waste: $70.12 COST ALLOCATION: RGDM (internal): $2,125.91 (26 sessions) SESSION QUALITY: HEALTHY Avg score: 91/100 | Cache hit: 79.1% | Flagged: 0/26 OVERHEAD TREND: STABLE Latest avg input: 172,046 tokens/session | WoW: -91.1% | Consecutive growth: 2 weeks RECOMMENDATIONS: 1. [HIGH] Route trivial sessions (<5 min) to Sonnet/Haiku instead of Opus (est. $280.48/mo savings) 2. [MEDIUM] Investigate low cache hit rate — may indicate context management issues ALERTS: [MEDIUM] (efficiency) Opus ratio 100% exceeds 25% threshold [MEDIUM] (efficiency) Estimated $70.12 waste from 10 trivial Opus sessions [LOW] (allocator) RGDM (internal) accounts for 100% of AI spend — consider cost optimization OUTCOME TRACKING: [=] Opus Ratio: 100% (was 100%, +0.0% — unchanged) [-] Est. Waste: $70.12 (was $0.58, +11989.7% — regressed) [-] Budget Pacing: 21% (was 21%, +0.5% — regressed) [+] Avg Quality: 91/100 (was 89/100, +1.6% — improved) [+] Cache Hit: 79.1% (was 76.9%, +2.9% — improved) [+] Overhead: 172,046 tok (was 1,098,002 tok, -84.3% — improved) 4-week trends: Budget Pacing: 71% → 17% → 19% → 21% (improving) Opus Ratio: 100% → 100% → 100% → 100% (worsening)
- › Route trivial sessions (<5 min) to Sonnet/Haiku instead of Opus
- › Investigate low cache hit rate — may indicate context management issues
MEDIUM
portfolio-news-monitor
6d ago
HITI THESIS EVENT: DOJ/DEA immediately placed FDA-approved + state-licensed medical cannabis into Schedule III (April 24, 2026). DEA hearing for broader rescheduling begins June 29, 2026. Thesis pillar REGULATORY TAILWINDS confirmed/accelerated. SAFER Banking: passed House + Senate Banking Committee, floor vote still pending. No HITI company-specific news or insider filings today.
▾
HITI THESIS EVENT: DOJ/DEA immediately placed FDA-approved + state-licensed medical cannabis into Schedule III (April 24, 2026). DEA hearing for broader rescheduling begins June 29, 2026. Thesis pillar REGULATORY TAILWINDS confirmed/accelerated. SAFER Banking: passed House + Senate Banking Committee, floor vote still pending. No HITI company-specific news or insider filings today.
HITI THESIS EVENT: DOJ/DEA immediately placed FDA-approved + state-licensed medical cannabis into Schedule III (April 24, 2026). DEA hearing for broader rescheduling begins June 29, 2026. Thesis pillar REGULATORY TAILWINDS confirmed/accelerated. SAFER Banking: passed House + Senate Banking Committee, floor vote still pending. No HITI company-specific news or insider filings today.
- › Review RGDM regulatory_event trend over the past 7 days for pattern
- › Check if any recent changes (bids, budgets, targeting) correlate with this alert