AI and machine learning
OpenAI API
OpenAI models behind a provider abstraction, with structured outputs and an evaluation set so the feature stays correct.
How I use it
Bolt, a PDF interaction product, runs on the OpenAI API with Stripe billing and Kinde auth around it. I put the model behind an interface so it can be swapped or given a fallback, use structured outputs where the response feeds code, and measure every prompt change against a fixed set of real questions.
Offered through
Shipped with OpenAI API
All projects →
Dubbix
Cross-platform AI-powered audio and video processing app with Text-to-Speech, Speech-to-Speech, Accent Changer, Video Dubbing, and YouTube Summarizer
Viral Shorts Agent with N8N
Automated workflow using n8n that fetches trending topics, generates viral short videos, and publishes them continuously on multiple social platforms
Bolt - Revolutionary PDF Interaction Platform
A conversational AI assistant built with OpenAI's GPT model to talk to PDFs
OGE (Operational Gaming Environment)
Game-based learning platform for professors and students with AI-guided scenarios and real-time collaboration
Questions clients ask about OpenAI API
- OpenAI or Anthropic?
- Both are good and I build against both. The choice is a one-line change if the system is designed for it, which is how I design it.
- How do you control cost?
- Smaller models where they pass the evaluation set, caching of stable prompt prefixes, and measuring tokens per completed task rather than per request.
Work with me
Need OpenAI API work done?
I work remotely with clients across the US, Canada, the UK, the UAE, Brazil and Australia. Tell me what you are building and I will reply with questions, a plan and an estimate.
- Published in IEEE Access, 2025
- BSc CS, Gold Medalist
- Clients in 6 countries
- 10 shipped projects