I have extensive experience in language modelling and the newest NLP technologies such as transformers and GPT. I have worked on web-scale language models for Microsoft for Windows and the search engine Bing. I also enjoyed doing research on NLP techniques and building machine learning pipelines based on transformers at Huawei, where I filed 3 NLP patents. Moreover, I have introduced NLP techniques to startups including GPT text summarization, question generation, and question answering. At the moment, I am working on fine tuning large language models to compete in the Bittensor network https://bittensor.com/. I also have 7 AWS certifications.
I am looking for projects that involve challenging and innovative machine learning and NLP tasks. I can help you with:
- Building your machine learning MVP
- Doing data analyses
- Building machine learning pipelines
- Building fully functional machine learning products (frontend and backend)
- Architecting infrastructure
- Building teams of data scientists and machine learning engineers
If you are interested in working with me or learning more about my skills and experience you can find more at: https://simoneromano.com
We are building machine learning products for a fixed weekly fee. We saw that search engines are the most requested type of product.
Nowadays, with https://streamlit.io/, https://typesense.org/, or http://jina.ai/ (which are open source) and some coding skills you can build state-of-the-art quality search engines. 10 years ago a team of engineers was required for this.
Thanks for the question. Not now. It is a beta version. I will definitely implement this if I see that people care about this service.
There are two things I plan to do: 1) typing the name of a company that automatically gets translated to a logo to match 2) type the name of a company that gets searched on the webpages to see if it is a partner
My initial idea was finding logos from landing pages to see what are the partners of a given startup. Before doing this I guess I have to get to a decent level of accuracy though.
I am just testing out some machine learning to extract tags from HTML. This is a very initial beta version. At the moment, I can find out tags with names, testimonials (e.g. they come under a section with title testimonials), and logos.
Of course it must be improved. But let me know anyway if you see some application for this tech.
I mainly created this because self promotion is frowned upon in many forums (e.g. HN, Reddit).
Here each post has two URLs: 1) The original forum where a question was asked 2) A product
Your product will be linked to the original discussion, and will get some visibility when indexed by search engines. This could also help to avoid self promotional posts on community forums.
I initially built if for me. I collected a DB of companies and their customers and wanted to sell it online. I couldn't really find an easy way to do it. If you are interested, I wrote about it here: https://medium.com/@ialuronico/how-to-sell-a-table-d7ec8d01d...
And regarding the price, yes it might be steep. I am open to revising it.
I was thinking that using the tools available now (e.g. Gumroad + Airtable, or Buy me a Coffee + Airtable) it could be competitive. They charge around a 5% fee for each transaction. If you sell a table for $50 they will become more expensive when you have more than 12 users.
If you build this out of no-code tools, e.g. Webflow + Memberstack, their subscriptions should add up to the same amount.
I have extensive experience in language modelling and the newest NLP technologies such as transformers and GPT. I have worked on web-scale language models for Microsoft for Windows and the search engine Bing. I also enjoyed doing research on NLP techniques and building machine learning pipelines based on transformers at Huawei, where I filed 3 NLP patents. Moreover, I have introduced NLP techniques to startups including GPT text summarization, question generation, and question answering. At the moment, I am working on fine tuning large language models to compete in the Bittensor network https://bittensor.com/. I also have 7 AWS certifications.
I am looking for projects that involve challenging and innovative machine learning and NLP tasks. I can help you with:
- Building your machine learning MVP - Doing data analyses - Building machine learning pipelines - Building fully functional machine learning products (frontend and backend) - Architecting infrastructure - Building teams of data scientists and machine learning engineers
If you are interested in working with me or learning more about my skills and experience you can find more at: https://simoneromano.com