API Management in the AI Era

The slide deck from the API Management in AI Era session from API Days Australia.

https://www.apidays.global/events/australia

Transcript

1. Nilesh Gule @nileshgule API Management in AI Era
2. $whoami { “name” : “Nilesh Gule”, “role” : “Senior Cloud Solutions Architect at Avanade” “website” : “https://www.HandsOnArchitect.com", “github” : “https://GitHub.com/NileshGule" “twitter” : “@nileshgule”, “linkedin” : “https://www.linkedin.com/in/nileshgule”, “YouTube” : “https://www.YouTube.com/@nilesh-gule” “likes” : “Technical Evangelism, Cricket”, }
3. Mistral Prompt and Response Tokens
4. Deepseek - Prompt and Response Tokens
5. API Management - GenAI Gateway Azure-Samples/AI-Gateway: APIM
6. Challenges in using GenAI APIs • Track Token usage across multiple applications • Ensure single app doesn’t consume whole TPM quota • Distribute load across multiple endpoints • Ensure committed capacity in PTUs is exhausted before falling back to PAYG instance
7. Monitor utilization of models • Sends Token Merics usage to Applications Insights • Provides overview of utilization of models across multiple applications or API consumers GenAI Gateway Capabilities in Azure API Management
8. Token Metrics Emitting – Token Metric Policy GenAI Gateway Capabilities in Azure API Management
9. Token Metrics Emitting GenAI Gateway Capabilities in Azure API Management
10. Enforce limits per consumer • Manage and enforce limits per API consumer based on the usage of API Tokens GenAI Gateway Capabilities in Azure API Management
11. Token Rate Limiting – Token Limit Policy GenAI Gateway Capabilities in Azure API Management
12. Token Rate Limiting GenAI Gateway Capabilities in Azure API Management
13. Provisioned Throughput Units (PTU) • Allows to specify the amount of throughput required in a model deployment. • Granted to subscription as quota • Quota is specific to region and defines the maximum number of PTUs that can be assigned to deployments in the subscription and region • PTU provides • Predictable performance • Allocated processing capacity • Cost savings Understanding costs associated with provisioned throughput units (PTU)
14. HA - Load Balanced Pool and Circuit Breaker • Helps to spread load across multiple Azure OpenAI endpoints • Round-robin, weighted or priority based load distribution strategy GenAI Gateway Capabilities in Azure API Management
15. Load Balanced Pool and Circuit Breaker GenAI Gateway Capabilities in Azure API Management
16. Load Balanced Pool and Circuit Breaker GenAI Gateway Capabilities in Azure API Management
17. AI Gateway capabilities of Azure API Management AI Gateway Security & safety • Keyless managed identities • AI Apps & Agents Authorizations -New • Content Safety -GA • Credential Manager Resiliency • Weight load balancing • Priority routing to provisioned capacity models • Backend pools with circuit breaker • Session aware load balancing -GA Scalability • Token rate limits and token quotas • Semantic Caching -GA • Model load balancing • Multi-regional deployments Traffic mediation & control • Azure AI Foundry & Azure OpenAI • OpenAI compatible models -GA • Responses API -GA • WebSocket’s for Realtime APIs • MCP server pass-trough - Soon • Expose APIs as built-in MCP server - Preview Developer velocity • Wizard policy configuration experience • Self-service with the Developer Portal • API Center Copilot Studio connector - Preview • Policy Toolkit Observability • Token counting per consumer • Prompts and completions logging -GA • Built-in reporting dashboard -GA Governance • Policy engine with custom expressions • API Center MCP server registry - Preview • Federated API Management GA GA GA Soon GA GA GA GA Preview Preview GA Preview New
18. Summary • Track Token usage across multiple applications • Emit Token Metrics policy • Ensure single app doesn’t consume whole TPM quota • Token Limit Policy • Distribute load across multiple endpoints • Backend pool load balancing and circuit breaker
19. Resources • Azure OpenAI Gateway topologies • Azure OpenAI Token Limit Policy • LLM Token Limit Policy • Azure OpenAI Emit Token Metric Policy • LLM Emit Token Metric Policy • Houssem Dellai Youtube videos • GenAI Labs • Designing and implementing GenAI gateway solution
20. Nilesh Gule ARCHITECT | MICROSOFT MVP “Code with Passion and Strive for Excellence” nileshgule @nileshgule Nilesh Gule NileshGule www.handsonarchitect.com https://www.youtube.com/@nilesh-gule
21. Source Code & slide deck Nilesh Gule fork - GenAI Labs https://github.com/NileshGule/AI-Gateway GenAI Labs https://aka.ms/apim/genai/labs https://speakerdeck.com/nileshgule/ https://www.slideshare.net/nileshgule/
22. Q&A