OpenAI API Platform from an RD perspective
Use Cases and Deployment Scope
We have utilized the OpenAI API platform to develop document data extraction and analysis capabilities. Also for agenting and chatbot applications as well as speech-to-text capabilities. So, I think you could say that we have used a very large scope of features from the API in our products and services.
Pros
- The developer experience is top notch. Their SDKs are super easy to use
- Organization and project billing separation. You know where everything was consumed.
- Playground. The playground is super useful to prototype without writing a single line of code
Cons
- I would like to be able to make structured outputs in prompts with pdfs included.
- The speech to text model (whisper) can be found in other inference providers like Groq for a cheaper price
- The rate limits can be a bit steep if you neeed to roll out quickly
Return on Investment
- The development time is pretty quick with their SDKs, so that's a positive ROI
- The whisper model can be found cheaper elsewhere (Groq for example), so although is not a negative ROI, it's a lost savings opportunity
- The quality of the models is great. So every product or service that uses them gets better, increasing our pricing power with customers
Usability
Alternatives Considered
Google Gemini and Anthropic Claude
Other Software Used
Google Gemini, Cursor, Supabase, Vercel, v0 by Vercel, Lovable, Interactive Brokers
