co-founder @ recyco & @lacasa_hq || prev @JaneStreetGroup || argentino 🇦🇷🌞ʕ•̫͡•ʕ•̫͡•ʔ•̫͡•ʔ•̫͡•ʕ•̫͡•ʔ•̫͡•ʕ•̫͡•ʕ•̫͡•ʔ

San Francisco, CA
Joined September 2015
Iñaki Arango 🦕 retweeted
Today Hone launches Engines: AI that owns business outcomes. Chatbots give answers. Agents do tasks. Engines deliver outcomes, and we hold them responsible for creating real value, measured in actual dollars instead of tokens or vibes. You point it at a business metric, and the Engine figures out the rest. It onboards itself, builds the necessary agents, skills, and memories, collaborates with the right people, and gets better with experience. From AI leaders to global enterprises, our customers have already used Engines to: • Uncover millions in procurement savings • Optimize product funnels • Manage credit risk • Increase qualified pipeline in new territories • Accelerate implementations and improve account management To scale this impact across the economy, we’ve raised a $60M seed led by Benchmark and Index with participation from Elad Gil, Hanabi, Definition, Diffusion, Lux, SV Angel, and Align. AI’s raw capabilities are far ahead of the value it creates in the real world, and Engines will help close the gap. Here’s how they work…
132
79
87
863
293,661
Iñaki Arango 🦕 retweeted
Simulating the consequences of an experiment before running it is a holy grail of AI for biology. How can we get there? Our new paper, “World models for biomedicine,” which I led together with @njwfish, is out now in @CellCellPress’s special issue on AI in biology. We explore how biomedical world models could help scientists navigate the enormous space of possible biological interventions. In principle, they could: 🧪 Simulate how biological systems respond to interventions never previously tested 🎯 Search for interventions that drive cells, tissues, or patients toward desired states 🧭 Plan experimental campaigns over long time horizons 🤖 Power autonomous, closed-loop AI scientists However, realizing this vision requires much more than calling a model a “world model.” What exactly defines a biomedical world model, and what capabilities distinguish it from related models? Where do today’s models already show promise, and where do they fall short? What new data and architectures are needed? What evaluations would demonstrate that a world model works? We tackle these and other questions in this latest paper from @marinkazitnik's lab: cell.com/cell/fulltext/S0092… More below 🧵👇🏽 1/8
22
55
9
281
18,922
Iñaki Arango 🦕 retweeted
a 10 minute application for a $50,000 fellowship. no strings. a total of $1,000,000 going out, towards making more magic in the world. presenting the Basis Fellowship. run by @markkhrapko ,@thewildstevenp , and spencer green
68
46
27
288
161,216
Iñaki Arango 🦕 retweeted
Introducing Instaplay: Make or Play Anything The world is burning trillions on AI to make us more productive - almost none of it has made life more fun. @InstaplayAI is changing that. 100,000+ users are already exploring what games can become when anyone can make one. Creators have already built multiplayer zombie FPS campaigns, interactive classroom lessons, and someone even proposed with a game. Our mission is to make AI the most powerful force for fun in the world: instaplay.ai
60
33
16
157
39,953
Iñaki Arango 🦕 retweeted
Introducing Aidan, your newest AI employee Aidan is the first computer-using AI built for realtime conversation We’ve raised $45M from @sequoia and @8VC to bring Aidan to the world @NotionHQ and @DecagonAI already use Aidan to run customer interactions. This is how @sablehq works:
299
143
138
1,540
742,922
Iñaki Arango 🦕 retweeted
Introducing Altara: the scientific intelligence platform for the physical world. Today @evatuecke and I are excited to announce our $7M seed led by @GreylockVC, joined by @Neo, @BoxGroup, @Liquid2V, and angel investors including @JeffDean and leadership from OpenAI & AMD. We’re already working with early customers in semiconductors, batteries, and advanced materials. More below.
135
133
34
863
191,793
Iñaki Arango 🦕 retweeted
What if you could trust your knowledge graph? We present OptimusKG, a multimodal biomedical KG built from high-quality structured and semi-structured resources to preserve factual, type-specific metadata across molecular, anatomical, clinical, and environmental domains. Highlights: Integrates 65 heterogeneous resources grounded with 18 ontologies and controlled vocabularies using the BioCypher framework and the Biolink Model. Contains 190,531 nodes across 10 entity types, 21,813,816 edges across 26 relation types, and 67,249,863 property instances encoding 110,276,843 values across 150 distinct property keys. Independently validated using PaperQA3, a multimodal agent that retrieves and reasons over scientific literature. Reproducible, deterministic and infrastructure-agnostic data pipeline with parallel execution. Distributed as Apache Parquet files, providing a standardized resource for graph-based machine learning, knowledge-grounded retrieval, and biomedical discovery. Built in collaboration with an interdisciplinary team of co-authors: @ayushnoori, @InakiArango, @JoacoPolo3, @SGRodriques, @andrewwhite01, David Clifton, @marinkazitnik To learn more about OptimusKG, here are some useful links about the release: Website: optimuskg.ai/ Documentation: optimuskg.ai/docs Code: github.com/mims-harvard/Opti… Paper: arxiv.org/abs/2604.27269 Client: pypi.org/project/optimuskg/ Dataset: doi.org/10.7910/DVN/IYNGEV This work was conducted across @Harvard, @harvardmed, @HarvardDBMI, @EdisonSci, @KempnerInst, @harvard_data, @UniofOxford, @broadinstitute
6
15
1
69
8,722
Iñaki Arango 🦕 retweeted
Better AI is not only about fetching more facts. It is about exploring and knowing when to switch gears mid-thought arxiv.org/abs/2601.13969 LLMs excel at language, but much of the world’s knowledge lives outside public web text: Internal ontologies, curated catalogs, drug safety tables, patient personal health data, lab knowledge bases Most systems force a choice: either search broadly like text retrieval or commit early to multi-hop exploratory traversal ARK is an agentic retriever that lets the model decide, step by step, when to zoom out for coverage and when to dive deep into relations No fragile seed selection. No fixed hop limits. No retriever training. Just two tools: 🔍 global search to surface relevant entities anywhere 🧭 local graph exploration to follow relations when evidence is multi-hop Stellar work by Joaquín Polonuer @lvvittor @ayushnoori @InakiArango @HarvardDBMI @harvardmed @broadinstitute @KempnerInst @harvard_data
2
19
2
108
7,564
Iñaki Arango 🦕 retweeted
I’m excited to share PROTON, a pre-trained relational transformer for neurology. 🧠 PROTON generates testable neurological hypotheses, which we evaluated across: 🧬 Parkinson’s disease: genome-wide in vitro and in vivo screens 🧫 Bipolar disorder: patient-derived brain organoids 🏥 Alzheimer’s disease: 610,524 patients tracked across 7 years By pairing AI with real-world experiments, PROTON defines a path for AI-driven discovery in neurological disease. Paper: arxiv.org/abs/2512.13724 Website: protonmodel.ai Code: github.com/mims-harvard/PROT… Model: huggingface.co/mims-harvard/… Learn more below 👇🏽 [1/8]
17
38
3
239
24,715
Iñaki Arango 🦕 retweeted
Replying to @NeurIPSConf
@NeurIPSConf week Sat, Dec 6 🔹 CUREBench - International Competition on AI agents and reasoning models for therapeutics at scale curebench.ai @GaoShanghua @ScientistTools @RichardYXZhu @sui67713 @ZKong50693 @xiaorui_su with a keynote by @AziziShekoofeh Sun, Dec 7 🔹 6th AI for Science NeurIPS Workshop – The Reach and Limits of AI for Scientific Discovery ai4sciencecommunity.github.i… @AdaFang_ We started this workshop series in 2021, when @AI_for_Science was still niche. It is exciting to see how the community and the field have grown and how much potential there is to transform scientific discovery Sat, Dec 6 🔹 AI Virtual Cells and Instruments – A New Era in Drug Discovery and Development ai4d3.github.io/2025/index.h… @_michellemli We are also presenting many papers throughout the week. Here is the first batch and more to follow throughout the week: • A scalable data layer of knowledge graph AI: openreview.net/pdf?id=8OXD0y… by Lucas Vittor, @ayushnoori @InakiArango, Joaquin Polonuer • Multi-agent collaboration in knowledge graph environments: openreview.net/pdf?id=xUDGCh… @ayushnoori @InakiArango Lucas Vittor, Joaquin Polonuer • Evolutionary reasoning in protein language models: openreview.net/pdf?id=nR9S0I… @YEktefaie Kudos to all stellar students and many thanks to fantastic collaborators @HarvardDBMI @harvardmed @Harvard @broadinstitute @KempnerInst
1
11
2
36
5,205
Iñaki Arango 🦕 retweeted
I've spent my whole life trying to make absence mean something. When I was five, the piano bench became my world. Practicing eight, ten hours a day wasn't punishment; I learned early to speak love through sacrifice. Every practice session was also a choice to say no. No to field trips, no to sleepovers, no high school prom, no summer trips, no inside jokes that last decades. By the time I realized what I'd traded, those years were gone. Missing became a habit. A personality. People learned not to expect me; I learned not to expect to be included. But here's the contradiction I can't shake: the thing that motivates me most is human connection. I want to be the person you call first – for your wins and your heartbreaks. I want to love people in the currency of time. With my presence. But I've built a life that keeps me in debt. A few days ago, I missed my little brother's 18th birthday. I didn't even notice what day it was. I know if I died today, the version of me most people know - always working, always "too busy," always absent - would be the opposite of who I am inside. So I made a deal with myself: if I can't buy back the time I've lost, maybe I can help buy it back for the hundreds, thousands, millions of others trapped in the same cycle. Maybe I can help them reclaim their presence – their ability to breathe, to connect, to be there for the people who matter before it's too late. That mission is Dex, a Chrome-based productivity tool designed to remember, understand, and save you time so you have energy for what matters most. To my friends and family: I think about you all the time. You don't know that, because thoughts don't hug back. I'm sorry for being a ghost. I'm building something that protects your time, and I'm working to protect mine with you. We can't get back the years we lost, but we built Dex to give you yours. Dex is available now.
150
72
45
669
251,752
Iñaki Arango 🦕 retweeted
Puentes is back. We’re flying out 10 engineers from Latin America for 7 days to… > eat dinner with @rauchg, @mejiasebas, and @tnm > get hired at top sf startups > move to San Francisco reply and I’ll dm you a link to apply
39
56
32
238
60,985
Iñaki Arango 🦕 retweeted
The DHS order hurts not just current Harvard students but grads too. Even years after college, grads on OPT visas may get kicked out of the US. If your status is in jeopardy, @Neo can help you join a startup or start your own, get an O-1, and stay in the US. A few options: 🧵
6
28
5
115
26,011
Iñaki Arango 🦕 retweeted
hiring interns for the summer to build cutting-edge neurotech with us!! we're agnostic to skillset — you just have to be amazing, and we'll take it from there. specific requirements: orbit.engineering/join
11
14
6
125
68,698
Iñaki Arango 🦕 retweeted
Puentes is starting this Monday! I'm flying 8 engineers to a mansion in SF. Hosting dinners every night with founders like @rauchg, @fin465, and @tnm. My goal is to expose them to the city and create opportunities for them to stay. DM if you want to help! Meet the engineers👇🏼
19
15
12
213
53,412
Iñaki Arango 🦕 retweeted
We won 2nd place overall at @hackwithtrees with Supernova: an AI Ad-maker for TikTok. Check out an ad we made of @AravSrinivas talking about @Perplexity_AI’s Deep Research below: You go from a product link → Tiktok Ad with a real influencer avatar talking, realistic b-roll, music, captions, the whole nine yards. Here’s how it works: Product → Market Research: 1. Drop a product link from etsy, amazon, or any shopping page. 2. Supernova will deterministically extract a product profile with @firecrawl and @GroqInc's R1-distilled model) 3. Our agents crawl the web to gather and synthesize market research from articles, competitors, profiles, and reports (powered by Perplexity Sonar API, Groq, and @OpenAI). Influencer Avatars: 4. Choose from hundreds of top influencers with detailed social and psychographic profiles (audience, engagement, quotes, beliefs) sourced from TikTok, @FamousBirthdays, and public articles. 5. Supernova reasons, plans, and generates a detailed creative script for 30-second ads with A-roll and B-roll, informed by hundreds of pages of product and market research. 6. Using @heygen/@synclabs, we create a hyper-realistic influencer avatar and 30 seconds of that influencer speaking with lip, hand, and face movements matching dialogue. For influencers/celebs, we also clone and optimize voice using @cartesia voice to voice! B-Roll, Music, Captions: 7. Supernova’s @runwayml pipeline uses the Gen3Alpha image-to-video model to create b-roll clips, with and without the product, optimized through our deep market research. Our b-roll always delivers—see for yourself! 8. And finally, we use @ZapCapAI for captions and music. Supernova delivers data-driven AI ads, making influencer marketing easy for teams of any size. No more AI slop—just precise, impactful marketing. Wouldn't have been possible without the most cracked team: @pranavramesh25 @InakiArango @nikitamounier this was one for the books @stanford🌲
10
7
2
144
25,537
Iñaki Arango 🦕 retweeted
Super excited to share that @dineshtva, @InakiArango, @nikitamounier, and I won 2nd Place Grand Prize at @hackwithtrees this weekend. We built Supernova, where we create AI-based short form user generated content from just a product link. We're able to generate realistic, TikTok-style ads for the likes of @on_running , @perplexity_ai , and @nvidia in under 3 min. - A-roll is generated with @HeyGen_Official and @synclabs and B-roll is generated with @runwayml. - Voiceovers are generated with @cartesia. - Scripts are rigorously curated from market research gained using @perplexity_ai and our own custom pipelines. - Insights are quickly extracted and processed using @GroqInc. This was our most ambitious project yet, built in just 30 hours. Check it out below! youtube.com/watch?v=0lCofRTA… devpost.com/software/superno…
14
5
2
56
10,676
Iñaki Arango 🦕 retweeted
search algorithms are cool, but have you ever seen a maze being solved completely by a diffusion model?
8
10
157
20,946
Iñaki Arango 🦕 retweeted
If you’re a woman engineer, founder, or investor who is international, we’d love to host you for dinner at La Casa this month. Please reach out!
2
3
9
1,251
Iñaki Arango 🦕 retweeted
Loved organizing the first of many @lacasa_hq house dinners tonight with @InakiArango. Thanks @BainCapVC @a16z for supporting. Interested in attending the next one? Please reach out! Thanks to @kevinzhang, @slaterstich, @BenLevy0, @archikadogra, @jonathanlei0 1/2
2
7
57
5,307