API Documentation & AI

Talk introducing Superface Edgar for NLP assisted API integration.
https://apithedocs.org/virtual-2023/zdenek-nemec

Transcript

1. Zdeněk “Z” Němec, February ‘23 API Documentation & AI Augmented API docs & API integration superface.ai
2. Founder & CTO of Superface.ai • Pioneered the API- fi rst approach at Apiary.io → Oracle • Founder of API consulting Good API • Helped several Fortune 100 companies with their API strategy and execution
 • Founder and CTO of Superface.ai Zdeněk “Z” Němec superface.ai
3. SaaS API integrations fatigue superface.ai
4. APIs are hefty burden Accelerated by global pandemic, the need for API integrations grew in 80,4% of companies 700 days Average time needed by a SaaS app to build API integrations superface.ai
5. Why integrations takes so long? superface.ai
6. The truth about API integration Analysis Development Testing Deployment time budget Expectation Analysis Development Reality superface.ai
7. API Analysis superface.ai
8. Superface research Top 6 obstacles when integrating APIs 1. Understanding the API domain model and language 2. Answering the question “Can this API ful fi ll my use-case?” 3. The need to understand the entire ecosystem and documentation of API vendor in order to integrate even a small functionality 4. Ambiguity of my own product requirements 5. Quality of documentation, outdated API specs 6. Access hurdles, certi fi cation, API provider terms superface.ai
9. What are we analyzing? Implemenation Product Business Business rules, commercials, limits, regulations, certi fi cation, SLAs How to connect, API calls, JSONs, API Spec what API o ff ers, what capabilities, what use-cases it carters to superface.ai
10. API documentation today Technical by the (wrong) default • Most often, API docs focus on the technical aspects • The least important one, and the one that changes the most often • API docs are not focusing enough on the product’s capabilities and domain • Business side is usually under-documented Implemenation Product Business superface.ai
11. superface.ai
12. https://openbankinguk.github.io/dcr-docs-pub/v3.3/dynamic-client-registration.html#basics superface.ai
13. Minimizing the analysis today? The tools we have today • Improving the API documentation • Document the business and product aspects in addition to technical reference • SDKs, cURL, Postman collections, examples • Standardization & harmonization • Uni fi ed APIs • “APIs in front of APIs” • Nylas, Hyperswitch, Metapack, Merge superface.ai
14. Next-gen API analysis superface.ai
15. Next-gen API analysis • Analysis will be always needed (for non-commodities) NLP-based 
 augmented human operator Employing LLM/AI to assist in the proces of analyzing and integrating API NLP – Natural Language Processing
 LLM - Large Language Model Self-integrating apps 
 machine-to-machine communication Enabling applications to autonomously discover & connect APIs superface.ai
16. NLP & APIs superface.ai
17. AI client code-gen GitHub Copilot superface.ai
18. AI API interaction design Jurassic-X superface.ai
19. AI API integration OpenAI GPT-3 / ChatGPT superface.ai
20. Superface EDGAR integration code
 on demand analyze API API Your App direct API call EDGAR superface.ai “How to ful fi ll order?" +
21. Superface EDGAR superface.ai
22. AI-enhanced API docs superface.ai
23. API documentation + NLP superface.ai
24. superface.ai API documentation + NLP
25. API documentation + NLP superface.ai
26. superface.ai How AI change API docs?
27. …when it doesn’t need the technical parts? Implemenation Product Business superface.ai
28. • Documentation of business aspects 
 • Explanation of domain models 
 • Focus on use-cases superface.ai
29. Thank you! Let’s talk! Zdenek “Z” Nemec 
 Twitter: @zdne superface.ai superface.ai