Hey HN, we are Max, Kieran, and Aahel from Midship (<a href="https://midship.ai">https://midship.ai</a>). Midship makes it easy to extract data from unstructured documents like pdfs and images.<p>Here’s a video showing it in action: <a href="https://www.loom.com/share/ae43b6abfcc24e5b82c87104339f2625?sid=2f2cf0fa-d671-4590-992a-da51712c69e1" rel="nofollow">https://www.loom.com/share/ae43b6abfcc24e5b82c87104339f2625?...</a>, and a demo playground (no signup required!) to test it out: <a href="https://app.midship.ai/demo">https://app.midship.ai/demo</a><p>We started 5 months ago initially trying to make an AI natural language workflow builder that would be a simpler alternative to Zapier or Make.com. However, most of our users seemed to be much more interested in the basic (and not very good) document extraction feature we had. Seeing how people were spending hours a day manually extracting data from pdfs inspired us to build what has become Midship!<p>The problem is that despite all our progress in software, huge amounts of business data still lives in PDFs and images. Sure, you can OCR them, but getting clean, structured data out is still painful. Most existing tools just give you a blob of markdown - leaving you to figure out which parts matter and how they relate.<p>We've found that combining OCR with language models lets us do something more useful: extract specific fields and tables that users actually care about. The LLMs help correct OCR mistakes and understand context (like knowing that "Inv#" and "Invoice Number" mean the same thing).<p>We have two main kinds of users today, non-technical users that extract data via our web app and developers who use our extraction api. We were initially focused on the first one as they seemed like an underserved part of the market, but we’ve received a lot of interest from developers who face the same issues.<p>For pricing, we currently charge a monthly Saas fee per seat for the web app and a volume based pricing for the API.<p>We’re really excited to share what we’ve built so far and look forward to any feedback from the community!
Heres a real world use case, our company has moved our pension provider. This provider like the old one sucks at providing me with a good way to navigate through the 120 funds I can invest in.<p>I want to create something that can paginate through 12 pages of html, perform clicks, download pdf fund factsheet, extract data from this factsheet into excel or CSV. Can this help? What's the best way to deal with the initial task of automating webpage interactions systematically?
I would like a tool that converts x months of credit card bills into a csv (the txn table from across PDFs and pages in each PDF) or something very easily.
Can you speak to the accuracy, particularly of numerical value extraction, that you’re achieving? I have a use case for pulling tabular financial data out of PDFs and accuracy is our main concern with using AI for that type of task.
Congratulations on the launch! Its a crowded space but I think there is place for a good and accurate tool!<p>Tried the examples - they seem tailored for specific document types. I have two questions around that: (a) is their a "best-effort" extraction you can perform or plan to support if you don't know the document type? (b) do you plan to support extraction from academic papers, i.e., potentially multi-column, with images, tables that are either single column or span two columns, equations, etc.?
Congrats on the launch! Just some friendly advice: financial documents such as quarterly earnings are actually highly structured via xrbl. If you are positioning the company as an unstructured -> structured process, then using these types of financial documents is probably not a great example even though everybody seems to do it.
Saw your demo video. Are you focusing on the finance sector primarily? It is a challenging industry IMO, requiring high accuracy and has strict privacy/security bar. How do you address these concerns?<p>Curious what are the biggest complain from your users? Are they willing to manually auditing the numbers in the table, make sure the output is 1. accurate. 2. formatted in the table they expected.
Congrats on the launch!<p>I’m curious to hear more about your pivot from AI workflow builder to document parsing. I can see correlations there, but that original idea seems like a much larger opportunity than parsing PDFs to tables in what is an already very crowded space. What verticals did you find have this problem specifically that gave you enough conviction to pivot?
Honest question but how do you see your business being affected as foundational models improve? While I have massive complaints about them, Gemini + structured outputs is working remarkably well for this internally and it's only getting better. It's also an order of magnitude cheaper than anything I've seen commercially.
Saw reducto released benchmark related to your product: <a href="https://reducto.ai/blog/rd-tablebench">https://reducto.ai/blog/rd-tablebench</a>
Curious your take on the benchmark and how well midship performs
I may or may not be the target audience, but it may help you to know a "book demo" link instead of a pricing page in the primary nav is a good heuristic shortcut for me to decide I'm not the target audience.
Congrats on the launch! A quick search in the YC startup directory brought up 5-10 companies doing pretty much the same thing:<p>- <a href="https://www.ycombinator.com/companies/tableflow">https://www.ycombinator.com/companies/tableflow</a><p>- <a href="https://www.ycombinator.com/companies/reducto">https://www.ycombinator.com/companies/reducto</a><p>- <a href="https://www.ycombinator.com/companies/mindee">https://www.ycombinator.com/companies/mindee</a><p>- <a href="https://www.ycombinator.com/companies/omniai">https://www.ycombinator.com/companies/omniai</a><p>- <a href="https://www.ycombinator.com/companies/trellis">https://www.ycombinator.com/companies/trellis</a><p>At the same time, accurate document extraction is becoming a commodity with powerful VLMs. Are you planning to focus on a specific industry, or how do you plan to differentiate?