yigitkonur35

thx!

enjoy it!

imo an agent learns way more by watching the raw agentic flow than by reading some sanitized context dump. you get to see exactly where the last bot derailed and then patched itself. give that a shot—handing over a spotless doc feels fake anyway.

yeah, thanks for sharing your thoughts! the original idea was to reproduce the exact same session on other coding platforms using the jsonl schema of each, but after seeing how the microsoft ai engineers on copilot-cli handle session continuity (major respect to them), i switched to single-message compression and it's been working very well for me.

most of us are juggling between $20 chatgpt and $20 claude subscriptions, so being able to hop between them by copying stuff over properly is clutch. give it a shot and lmk what you think, or drop a pr if you’ve got tweaks!

i've heard the boom of 'vibe coded' apps seems like apple's just overwhelmed and responding by tightening up, which makes it harder for everyone, even legit projects yours. btw, product looks really nice - hope that you get a good traction.

shows how bad embeddings are in a practical way

finally someone has made something like n8n that’s easy to observe but also offers coding flexibility, i was just about to dive into making an agent with a prompt chain that generates workflows on n8n when i found you. if llms.txt becomes a bit more practical, it’s written simply enough to have the potential to go from prompt to agent. shadcn level simplicity for writing agent is needed. if the right decisions are made, this product may be the new toolkit of the market is looking for.

great stuff! nice to see that remotion is becoming more popular on such projects.

I really appreciate you sharing your hands-on experience with a real-world scenario. It's interesting how people unfamiliar with traditional OCR often doubt LLMs, but having worked with actual documents, I know how inefficient classic OCR methods can be. So these minor errors don't alarm me at all. Your use case sounds fascinating - I might just incorporate it into my own benchmarks. Thanks again for your insightful comment!

You're right. For my API that prepares PDFs for LLMs, fixing typos makes sense. But yeah, keeping original text is crucial for most OCR tasks.

I've found this method really useful for prepping PDFs before running them through AI. I mix it with traditional OCR for a hybrid approach. It's a game-changer for getting info from tricky pages. Sure, you wouldn't bet the farm on it for a big, official project, but it's still pretty solid. If you're willing to spend a bit more, you can use extra prompts to check for any context skips. It's a lot of work, though - probably best left to companies that specialize in this stuff.

I've been testing it out on pitch decks made in Figma and saved as JPGs. Surprisingly, the LLM OCR outperformed top dogs like SolidDocuments and PDFtron. Since I'm mainly after getting good context for the LLM from PDFs, I've been using this hybrid setup, bringing in the LLM OCR for pages that need it. In my book, this API is perfect for these kinds of situations.

You're absolutely right. I use PDFTron (through CloudCovert) for full document OCR, but for pages with fewer than 100 characters, I switch to this API. It's a great combo – I get the solid OCR performance of SolidDocument for most content, but I can also handle tricky stuff like stats, old-fashioned text, or handwriting that regular OCR struggles with. That's why I added page numbers upfront.

I ought to test this with Sonnet too and compare the results. I feel it might perform better on OCR tasks. While I went with Azure OpenAI due to fewer rate restrictions, you've got a point - Sonnet could really shine here.

Wow, you knocked it out of the park! I'll be sure to use this when I tackle that evaluation.

You're spot on. We shouldn't lump all LLMs together. This approach might work wonders for Anthropic and OpenAI's top-tier models, but it could fall flat with smaller, less complex ones.

I purposely set the temperature to 0.1, thinking the LLM might need a little wiggle room when whipping up those markdown tables. You know, just enough leeway to get creative if needed.

Yes, you can customize this as you wish by adding it to your prompt.

I did a ton of Googling before writing this code, but I couldn't find you guys anywhere. If I had, I'd have definitely used your stuff. You might want to think about running some small-scale Google Ads campaigns. They could be especially effective if you target people searching for both LLM and OCR together. Great product, congratz!

For highly consistent responses, manually transcribing the most challenging page of the document (or engaging in multiple rounds of dialogue with Claude) and incorporating it as a few-shot example can dramatically improve overall consistency.

I get your worries about LLMs and their consistency problems. But I think we can fix a lot of that using LLMs themselves for checks. If you're after top-notch accuracy, you could throw in another prompt, add some visual and text input, and double-check that nothing's lost in translation. The cheaper models are actually great for this kind of quality control. LLMs have come a long way since they first showed up, and I reckon they've stepped up their game enough to shake off that bad rap for giving mixed signals.

You've got a point, but try testing it on a tricky example like the Apollo 17 document - you know, with those sideways tables and old-school writing. You'll see all three non-AI services totally bomb. Now, if you tweak it to batch = 1 instead of 10, you'll notice there's hardly any made-up stuff. When you dial down the temperature close to zero, it's super unlikely to see hallucinations with limited context. At worst, you might get some skipped bits, but that's not a dealbreaker for folks looking to feed PDFs into AI systems. Let's face it, regular OCR already messes up so much that...

I messed around with some rotating tables in that Apollo 17 demo video - you can check it out in the repo if you want. It's pretty straightforward to tweak just by changing the prompt. You can customize that prompt section in the code to fit whatever you need.

Oh, and if you throw in a line about LaTeX, it'll make things even more consistent. Just add it to that markdown definition part I set up. Honestly, it'll probably work pretty well as is - should be way better than those clunky old OCR systems.

People are really freaked out about hallucinations, but you can totally tackle that with solid prompts. The one in the repo right now is doing a pretty good job. Keep in mind though, this project is all about maxing out context for LLMs in products that need PDF input.

We're not talking about some hardcore archiving system for the Library of Congress here. The goal is to boost consistency whenever you're feeding PDF context into an LLM-powered tool. Appreciate the feedback, I'll be sure to add that in.

It is a lot cheaper! While cost-effectiveness may not be the primary advantage, this solution offers superior accuracy and consistency. Key benefits include precise table generation and output in easily editable markdown format.

Let's make some numbers game:

- Average token usage per image: ~1200 - Total tokens per page (including prompt): ~1500 - [GPT4o] Input token cost: $5 per million tokens - [GPT4o] Output token cost: $15 per million tokens

For 1000 documents: - Estimated total cost: $15

This represents excellent value considering the consistency and flexibility provided. For further cost optimization, consider:

1. Utilizing GPT4 mini: Reduces cost to approximately $8 per 1000 documents 2. Implementing batch API: Further reduces cost to around $4 per 1000 documents

I think it offers an optimal balance of affordability & reliability.

PS: One of the most affordable solution on market, cloudconvert charges ~30$ for 1K document (pdftron mode required 4 credits)

You should try Exa.ai API

Randomly Explains Ambiguous Development Methodologies Endlessly

Curious about Clickhouse’s approach to this compression structure.

There are few things that grind my gears more than hyping up a product only to not find it when I check the panel. I've been eagerly awaiting this one: https://upload.wope.com/R4nhkLdh

Anyone know if we’ll be able to use Google Cloud credits for this?

Sounds like a great value for those who don’t want to involve with k8s. As you integrate with Hetzner and their new billing system (not charging monthly anymore) is making your service more useful.

I'm out of options and tired of receiving 500-word cover letters within 2 minutes of posting job ads. It seems writing shorter, carefree cover letters might stand out more. Moreover, the AI that writes these cover letters isn't well-prompted, as it can go astray even with the most basic instructions.

I'm not personally bothered by the criticism of 'this style of design' or copywriting; rather, I find it troubling that such a reaction is given to a project that many people have worked on for weeks. Moreover, despite the fact that the design was made almost 2 years ago, we were very upset that it was described as if it was a mere imitation of a style that has become popular. Otherwise, of course, we can respect everyone's opinion on any subject.