Sentient Labs

We’re excited to announce our official partnership with alphaXiv to advance open-source research together. alphaXiv is one of the leading research platforms making it easier than ever to discover and parse through cutting-edge work in an era where AI research is progressing at a blistering pace, ushering in a new age of open academic discovery. Together, we’ll host ongoing community talks and spaces around open-source research, model security, and agent architecture, helping builders learn directly from the researchers and engineers pushing the frontier forward. Also, the alphaXiv community will be directly involved in bringing their research ideas to life through Sentient’s open-source foundation. As alphaXiv evolves into a true hub for global research, where you don’t just read papers, but also leverage datasets, benchmarks, and run easily reproducible experiments—we’re excited to partner with them and accelerate open-source AI development. Our shared goal is simple: make open research the fastest path to real progress. Open-source AI 🤝 open research.

⚡ Co-Founder of Sentient Labs, Professor of ECE at the Indian Institute of Science, Himanshu Tyagi joins the 15th anniversary Blockchain Life in Dubai 200+ industry leaders across 3 stages share insights with 15,000+ attendees worldwide. 🎟 Buy tickets https://lnkd.in/e8XrUvjk

Andrew Miller is one of the coauthors of the OML paper. He works at Teleport on applications of confidential compute in social media and communications. He holds a PhD from the University of Maryland’s  Cybersecurity Center and is an Adjunct Associate Professor at the University of Illinois Urbana-Champaign as well as an Associate Director of IC3 His expert background in cryptography and confidential computing make him an ideal collaborator on mechanisms that allow open-source AI artifacts to be open, monetizable, and loyal (OML). Andrew contributed to the section on confidential compute and TEEs. Alongside Andrew, we’ve co-authored and released a fundamental manuscript detailing the next steps toward OML, pushing the frontier of cryptography research powering open-source AI; see  https://lnkd.in/ehwumFFW for details. This work will be presented at the 2025 NeurIPS Conference during the Workshop on Lock-LLM.

We’re proud to congratulate our co-founder Sandeep Nailwal on being named to The Economic Times 40 Under 40 👏 We’re pushing the limits of what’s possible with open-source AI, keeping AGI free from single-entity control and open for everyone. Onward.

We were recently recognized as AI Startup of the Year at the Minsky Awards, presented by AIM, Leander Paes, and Rajgopal A S On the FRAMES benchmark for multi-step reasoning, ROMA Search achieved an industry-leading 81.7% accuracy, significantly surpassing competitors. While on the SEAL-0 benchmark for complex, multi-source reasoning, ROMA scored a 45.6% accuracy, also surpassing competitors. ROMA is available on GitHub and free to try for everyone, let us know what you're building!👇

**Join Sentient AI research team to build complex multiagent reasoning AI** Last month we released ROMA (https://lnkd.in/gPNZFpg2), the state-of-the-art framework for multiagent reasoning. It provides off-the-shelf planning, coordination and parallel execution for multiple agents and tools to complete complex AI tasks or respond to queries. It was (perhaps still is) the leader on difficult reasoning benchmarks like FRAMES, SEAL-0, etc. We are growing this framework to include options for faster reasoning architecture, support for more functionalities, improve search tool, and a self-improvement pipeline. This is at the heart of the reasoning AI stack from Sentient that allows you to build powerful agents (including the ones have put up on Sentient Chat). I am looking for AI researchers and engineers who can contribute to this framework. I am open to both full-time and internship roles. Write to me if you have done one or more of the following: 1. You have completed a masters degree or have other experience building agents using ReAct, etc. 2. You have published research on reasoning AI (for maths, coding, science, games, etc.) at top venues 3. You have setup eval framework for AI and/or AI improvement pipelines (prompt optimisation, post-training) 4. You have built AI apps that handle both structured and unstructured data 5. You have worked on AI search frameworks (like ODS from Sentient https://lnkd.in/gaweg6BG) I promise we will work on the most interesting problems and build the most interesting AI things. Apply here: https://lnkd.in/gtMJZCCz

The GRID continues to improve, with our newest upgrade coming to Sentient Chat. GRID is the world’s largest network of intelligence, containing agents, models, data sources, frameworks, and Sentient Chat—the infrastructure that stitches it all together. We’re proud of our GRID products—ROMA, ODS, Dobby, Fingerprinting, and Sentient Chat—and excited for users to build on top of them openly!

Very excited to release our newest piece of research: ROMA (Recursive Open Meta Agent). This framework outperforms every platform that does deep research, and breaks down complex queries into atomizable, achievable tasks. We're now #1 on GitHub trending repo and #11 on Product Hunt, really excited to see what people build on top of our work. For builders: ROMA makes multi-agent systems modular and extensible. Plug in different LLMs, tools, or prompts at any node. Run subtasks in parallel. Create agents for research reports, financial analysis, creative generation, and more. For researchers: ROMA is a transparent backbone for studying next-generation agent systems. Stage tracing exposes reasoning steps, making it easier to refine prompts, test strategies, and push the field forward. All the links 👉 Read our technical blog: https://lnkd.in/eZ5K7Zgh 👉 Develop with our fully open repo: https://lnkd.in/efZkGjUN 👉 Watch the announcement talk at SBC 2025: https://lnkd.in/earvZPT8

The engineering team is getting it done😁 We’re pushing hard with whole new Sentient Chat update. The team is excited to bring it live very, very soon.

Join us this Monday at the Open AGI Symposium at University of California, Berkeley, where Stanford University researchers John Yang and Mike Merrill will present work pushing the boundaries of open-source AI. John will introduce SWE-smith, a framework for scaling data in software engineering agents. Mike will share Terminal-bench, a new benchmark for evaluating AI agents in terminal environments. Reserve your ticket here: https://lu.ma/a4ndq3da