We raised $75M in Series B funding led by @a16z, bringing our total funding to $108M. @reductoai is defining the new frontier for document intelligence, helping the world’s leading AI teams unlock data from their unstructured documents. We’re launching a new pay-as-you-go tier with 15k free credits, allowing any company to get started with Reducto. Read the full blog in the comments.

what happens in tulum stays in tulum (except the vlog)

Loved working with the @DrataHQ team on this case study! Got to share how the AI workflows they’re building with @reductoai fit into their bigger vision for the future of trust and security 🔐🔮

🔗 reducto.ai/blog/reducto-drat…

Drata serves more than 8,500 organizations, including a third of the Cloud 100, and is building toward a future where AI agents handle more of the daily work behind compliance. Reducto makes customer documents usable across those AI workflows, from extracting security questions to writing approved answers back into the original files. Hear from Lior Solomon, VP of Engineering, AI and Data, and Julien Melissas, who leads Drata’s AI Questionnaire Assistance team, on how they put Reducto to work. @DrataHQ case study in the comments 👇

with r-1, we've replaced a long pipeline of specialized models with one model that sees the whole page. that ultimately translates into fewer handoffs where context can get lost. we invited @ani_0203_ and @chuyishang to walk through a few of the edge cases they worked on, from watermarks and handwriting to figure summaries that use the surrounding page for context. really proud of what the team built here!

Building agents at @opennote_me and now @reductoai has repeatedly taught me that agent capability will continue to grow, but “you can just give it tools” is true in principle and not in practice. I spoke at @aiDotEngineer a few months back about how to build the right tools and guardrails so the agent experience serves the user the right way: piped.video/watch?v=jJQoVkd5…

we're taking over NYC this week! come say hi if you're at @aiDotEngineer 🗽 find us at booth 34 in the Empire Ballroom all 3 days 🔮

we’re adding a pricing calculator so you can see what your document pipeline could cost as you scale 🔮 try it with your own numbers: reducto.ai/pricing

🥂 Join Reducto, Modal, Browserbase, and Mintlify for an Engineering x AI Leaders Happy Hour. 🗓️ Tuesday, October 13 · 6–8 PM 📍 The Skylark Rooftop Bar RSVP: luma.com/reductoai-s3br

bro i didn’t see you at the esoteric pdf meetup at @reductoai shoutout @PabloAntonio @brenniepell

what does "good" document processing even mean? benchmarking in this space can mean a lot of things: > accuracy on complex docs > latency > throughput for scale we're not here to tell you what set of documents is a "definitive benchmark", because your own docs are the ones that matter. the more important part is actually deciding how you measure performance. @archanaahlawat and @jannalulu_ have been working a lot on building and tuning our in-house judge framework, and turns out there's a couple of reoccurring patterns we see that might help you with your eval process:

tomorrow, at the @reductoai office ;) dm me for rsvp link

Benchmark Spotlight: @reductoai Benchmark: LegalParseQA (Document Processing, redlined contracts) Metric: Parse time (median and p95) The benchmark tests how well parsers handle heavily redlined contracts so an agent can tell what was struck, inserted, and still applies.

Next in our series with the team behind Reducto, @hu_yifei from the ML team shares how we balance accuracy, latency, and cost with r-1. He walks through making our smaller models more capable and routing pages based on complexity, so straightforward documents get a fast response while harder ones benefit from our larger models 🔮 🔗 Learn more about r-1: reducto.ai/blog/parse-r-1-mo…

not @reductoai raising the bar for startup offsites by renting out pablo escobar's entire old mansion 😭