OthmaneHamzaoui

But can't you still achieve domain-specific or privacy conscious cases by using retrieval + vector dbs?

Thanks for the feedback ! I Intentionally avoided the classic list/filter to provide more guidance through the questionnaire but I can see how that can limit the browsing experience (having to go back and forth), will explore adding a classic catalogue as an option :)

I'm less surprised by the acquisition but more confused by seeing MosaicML labeled as 'Generative AI startup'. To my knowledge MosaicML is an open source lib for optimising training/inference taking a "speedup" tricks from research papers and making them available easily. At least a few months back that was the case. When did they become a Gen AI startup ? Or they rebranded lately?

Pretty straightforward to use !

Maybe move the log-in page after step 2, seems more natural as it's the last step before seeing the results. Was a bit confused with the style component. It only changes the text font and not the style of the website. Either specify it's changing font only or generate different templates when a person selects a different style (playful, simple etc) :)

Also given your generate the template once I would show again the color/font changes to step 3 so users can interact with the page style after it's generated.

Makes the experience more engaging !

Interesting, happy to chat and provide feedback as I have been working in this field for the last few years. Did you get inspiration by any chance from the following paper : https://arxiv.org/pdf/2205.07147.pdf and their recent implementation https://github.com/skypilot-org/skypilot ?

On point ! Been working on putting AI/ML systems in production for various large companies in the last 5 years and every time the ML part of the system is just the tip of the iceberg. System integrations and user adoption are two main big components that had to be tackled before the AI system was in production.

I think the natural excitement we have with any new AI model (or technology) leads us to assume that it will magically get integrated in any existing system and adopted by any user.

For text based open-source models I would start looking in the HuggingFace Models hub (https://huggingface.co/models) and filter on "Feature extraction". Here is an example (https://huggingface.co/sentence-transformers/all-MiniLM-L6-v...). What do you plan to do with the embeddings once created?

For hosting solutions that depends on the traffic you're expecting. Is this just for testing or you have a level of traffic in mind?

Depends on what your work is. I would go ahead and assume you have the possibility to block time and decide what task you work on and when during the day. If not the below tips may not be applicable 100%.

These are based on my work experience in IT in a large and fast-paced FANG company.

1/Reduce the quantity: Push back against as much tasks as you can. "Urgent" tasks are rarely urgent..., take with a grain of salt. Use the "Delete, Delegate, Delay and Do" concept as a mental model if you're unsure (https://betterhumans.pub/delete-delegate-delay-do-8ab729ffd3...).

2/Use the "Most Important Tasks"concept. Pick one or two main things you need to get done each day and focus on that. Leave the rest for the end of day and check point 1/.

3/Finally look at time management/optimisation not just at how you optimise time today/tomorrow but also long-term. For example doing a task yourself might be more quick than training someone you work to do it. However once you train that person you can delegate it and get "time optimisation returns" in the future :)

Hope this helps, cheers !

Love the idea of connecting niche expertise wherever there is a market for it. Adopting technologies (even saas or low code) requires expertise sometimes and this could enable companies with limited bandwidth/talent. Reminds me of companies who provide experts networks.

The part where you extended your scope from lowcode to broader set of software is clear, didn't get the part about being an enabler instead of the platform it self, the network aspect is not clear either.

Website is not straightforward on mobile and images within the one liner text can be confusing. Your description on HN is much more clear:)

Question : How do you differiante from platforms like fiver ? different audience of expertise ?

I was able to add buy orders of random stocks, a loading window shows and disappears but then the overview/graph/metrics/investment segments keep being blured in and out non stop :)

As mentioned by gregjor business pay people to solve business problems but given your interest in pure tech I would look for companies who's core business is something purely technical (example: DevTool companies are a great place to start).

Unsure what you mean by "senior" roles, but in most large companies it's what you described as not wanting to do: tech + non tech responsibilities. If by seniority you mean owning a larger piece of the codebase and solving challenging problems I would look more for roles of "founding engineers" or simply "software engineer" at early stage startups (again building software that has a technical edge).

The smaller the organisation and the more its business is purely a tech product in itself the more likely it will align with what you're looking for (+ you'll be surrounded with like minded engineers). Hope it helps :)

Great tool, I just signed up, was only able to test the simulation feature (portfolio not working).

I like the simplicity of choosing Stocks/ETF or crypto. Would be great if you can select which ETFs are chosen in the portfolio (vs having automatic assignment).

The list metrics (ex: Holding 2 years ago) shown right after the simulation can use a "i" button explaining what each metric means.

Is the simulation based on "buying 2 years ago and not reselling" or using some kind of DCA?