I connect language models to things that already exist: a Telegram bot, a website form, an internal tool, a spreadsheet that needs text written or classified.
Typical jobs: a support bot that answers from your own documents instead of making things up, incoming messages or tickets sorted and routed automatically, product descriptions or summaries generated in bulk from a table, transcripts turned into structured notes, a first-line assistant that hands over to a human when it is out of its depth.
The part people underestimate is everything around the API call — keeping the cost per request sane, retrying properly when the API times out, not sending the whole database into the prompt, and making sure the output comes back in a shape your code can rely on. That is most of the actual work and it is what I do here.
Stack: Python, OpenAI and Anthropic APIs, FastAPI, aiogram, PostgreSQL, and pgvector or a small vector store when the bot has to search your own documents.
You pay the model provider directly with your own key, so nothing runs through my account and you keep control of the spend. Tell me what you want it to do and I will tell you whether it is a base-price job and what it will cost you per month to run.