Observability & Evaluation · Software component

Optimization Recommender

Software componentObservability & EvaluationObservability & Evaluationarc:OptimizationRecommender

An analysis component that turns agent profiling data into ranked remediation suggestions, such as parallelizing independent tool calls or caching repeated calls, each with estimated latency and cost impact.

Responsibility. Quantifies and prioritizes candidate workflow optimizations from profile data.

Also known as: Recommendation engine, Automated optimization recommendations, Caching opportunity analysis, Token distribution analysis, Cost optimization recommendations

Direct neighbourhood (hover for relationship types)

Relationships

reads dependency

sends data to dynamic

Agent Developer Ch7.3

Design guidance

SHOULD recommend parallelization when sequential tool calls share no data dependencies, with an estimated latency reduction.
SHOULD recommend caching when identical tool calls repeat, with projected cost savings.
SHOULD prioritize high-impact, low-effort optimizations before complex refactors such as prompt rewrites for token reduction.
SHOULD recommend caching only for content that is both frequently reused and large enough to justify caching overhead.
SHOULD analyze the application's specific input/output token distribution first: prioritize output reduction for low-volume short-prompt agents and input reduction for high-volume agents with large static context.

Quantitative guidance

As stated by the sources; verify before use.

Example recommendations: parallelize arxiv_search and database_query to cut latency ~40%; cache vector_search to remove 15 redundant calls saving $0.23 per transaction (Ch7.3).
Portfolio-recommendation analysis: 50,000 requests/month, $7,500/month, avg 9,000 input tokens (800 system prompt, 5,000 client profile, 3,000 market data, 200 query) and 800 output tokens; 5,800 cacheable tokens; $2,175/month projected caching savings (Ch8.3).
Weekly review recommendations: cache more RAG responses, optimize a top customer's prompts, consider model downgrade for simple queries (Ref8.05).

Classification

Patterns: Profile-guided optimizationImpact/effort prioritization
Technologies: NVIDIA NeMo Agent ToolkitNVIDIA Agent Intelligence Toolkit (AIQ)
Quality attributes: Performance efficiency (ISO/IEC 25010)Cost efficiency
Risks mitigated: Guesswork optimization (prompt/model/temperature changes without knowing the bottleneck)

Sources

Ch7.3: T. Nguyen, "NeMo Agent Toolkit Profiling," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 7.3. ISBN: 9798244538229.
Ch8.3: T. Nguyen, "Token Economics and Architecture," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 8.3. ISBN: 9798244538229.
Ref1.01: NVIDIA, "NVIDIA NeMo Agent Toolkit overview," NVIDIA NeMo Agent Toolkit Documentation, v1.8. Accessed: Sep. 26, 2026. [Online]. Available: https://docs.nvidia.com/nemo/agent-toolkit/latest/index.html
Ref8.05: "Cost Optimization and Resource Monitoring for Agent Systems," unpublished reference note (05-Cost-Optimization-Resource-Monitoring.md), Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam supplementary materials, 2026. unpublished note