Viewing profile — giordanol

[dead]

Hey HN! I built Notte to see if converting DOM into natural language could improve web agent capabilities and make them work reliably with smaller models. The result was using deep…

Bamboo growth pattern - years of invisible growth underground, then it suddenly shoots up 90 feet. Easy to forget how long the roots were forming.

Publishing early work feels pointless until you look back and realise the later stuff couldn't exist without it. Same goes for any expressive work. Sounds like a platitude, but it …

The tooling problem is 90% solved. The new technical bottleneck is human judgment.

Hi HN, This is Lucas, one of the creators of Notte. Notte is an open-source full-stack framework for web agents, designed to be fast and reliable in production. Our tech revolves a…

Feels like a semi-simple UX fix could make this a lot more natural. Git-style forks but for chats.

Would love to see metrics that isolate recovery behaviour (if any)

Agree with this. Constraining generation with physics, legality, or even tooling limits turns the model into a search-and-validate engine instead of a word predictor. Closer to pro…

Shift feels real. LLMs don't replace devs, but they do compress the value curve. The top 10% get even more leverage, and the bottom 50% become harder to justify. What worries me is…

Cursor’s doc indexing is acc one of the few AI coding features that feels like it saves time. Embedding full doc sites, deduping nav/header junk, then letting me reference @docs in…

LLMs shift the bottleneck - becomes less about typing code, more about spotting when something’s subtly wrong. Still need real judgment just applied to different layers. The skills…

Don't think the limit is in what LLMs can evaluate - given the right context, they’re good at assessing quality. The problem is what actually gets retrieved and surfaced in the fir…

Basically treating extraction as an adaptive loop instead of a static function. If first parse fails or looks incomplete, tweak the prompt, inject more context, or switch strategie…

Pretty cool. However truly reliable, scalable LLM systems will need structured, modular architectures, not just brute-force long prompts. Think agent architectures with memory, sta…

Value isn’t just the editor, it’s the workflow. Letting LLMs plan and act across multi-step flows is a hard problem, and Windsurf figured out a dev-focused version of that. Gains t…

Really cool direction. The embedding-first + agentic verification pipeline resonates, similar pattern worked well for us in the web interaction space.

Shift isn't just about competitors gaining ground but about users increasingly bypassing traditional search entirely. Between Reddit, Perplexity, ChatGPT, and direct domain knowled…

In my experience the key friction point has been schema stability vs input variance. Had better luck treating mapping as a dynamic planning problem with retries and memory.

Agreed. Catching mismatches between doc and implementation is still valuable, just wouldn’t want people to rely on it as a safety net when the docs themselves might be inaccurate/i…

A breakdown would be interesting. I can’t give you hard numbers, but in our case scaffolding was most of the work. Getting the model to act reliably meant building structured abstr…

Curious how you're handling multi-step flows or follow-ups, seems like thats where MCP could really shine especially compared to brittle CLI scripts. We've seen similar wins with b…

LLM-based coding only really works when wrapped in structured prompts, constrained outputs, external checks etc. The systems that work well aren’t just 'LLM take the wheel' archite…

Treating docstrings as the spec and asking an LLM to flag mismatches feels promising in theory but personally I'd b wary of overfitting to underspecified docs. Might be useful as a…