Software Engineer - Tools and Automation

Lausanne-based Adaptyv raises €34.35 million Series A to build the automated lab for agentic biology. August 27, 2026 Adaptyv Bio, the Lausanne-based BioTech startup building an automated wet lab for AI-driven protein design, has raised a €34.35 million ($40 million) Series A round. The round was led by Highland Europe, with existing investors Ace Ventures, ByFounders, and Y Combinator doubling down. Ace Ventures led the company's €6.8 million ($8 million) Seed round in late 2024. "AI has made it easy to design proteins. Many teams can now create thousands of candidate molecules for a target in an afternoon. Finding out whether those molecules actually work is still the hard part. That requires a wet lab. But most wet labs were built for a world where scientists tested twenty designs per target, not two thousand. Adaptyv closes this gap," ACE Ventures said. Founded by Julian Englert and Daniel Nakhaee-Zadeh Gutierrez, Adaptyv has built "an automated, AI-native lab" where the process starts with a digital protein sequence and ends with real experimental data. DNA is synthesised in-house, proteins are expressed cell-free, and binding is measured using SPR with built-in quality control. Customers can submit their designs through a web platform, or through an API, allowing AI agents to order their own experiments directly. When ACE led Adaptyv's Seed round in November 2024, the startup had only nine people and a lab that was beginning to see commercial demand. Fast forward to 2026: the team has tripled, revenue has grown roughly 10x, the company has brought DNA synthesis in-house, scaled lab automation, and even launched its Proteinbase competition platform, attracting "680 participants and 10,000 protein designs to one competition." Adaptyv has also shipped a public API now integrated with Boltz, Chai, Cradle, Tamarind, Latent Labs and Benchling. It now counts over 100 customers, including frontier AI labs, top 5 pharma companies and AI-native drug discovery startups. The company recently ran a benchmark with Anthropic, sending 16 targets from its public design competitions to Claude Science and testing the resulting sequences, anonymised. According to the company, 95% of the 1,320 designs were expressed. "Out of 1,320 designs, 354 bound their target, giving Claude an average hit rate of 26.8%. Only a single target yielded 0 binders. Compared to the results of our Protein Design Competitions, Claude's designs had noticeably higher success rates and would have won 5 out 6, yielding tighter binders," reported the company The test showcased how Adaptyv is "becoming the place frontier labs go when they need real-world proof". The company noted that Google DeepMind runs a similar validation loop, Chai Discovery used Adaptyv to validate its Chai-2 zero-shot antibody release, and Roche and Novo Nordisk run therapeutic design cycles roughly 4x faster than their internal pipelines. The new capital will fund two types of growth: horizontal expansion of lab capacity and vertical expansion into harder biology problems. On the horizontal side, the startup plans to roughly triple its lab capacity by the end of 2026 and keep growing through 2027. On the vertical side, it intends to move beyond binding measurements, which is a market estimated to be worth "a few hundred million dollars a year", into developability and full biophysical characterisation, next-generation modalities such as peptides, ADCs and degraders, and eventually cell-based function. For the horizontal expansion, Adaptyv Bio has announced a new lab and office will open in London in the fourth quarter of 2026, alongside an expansion in Lausanne. The company plans to grow its team from 25 to around 60, with a focus on production, automation and software. For its founders, the long-term goal is to build a biological gigafactory with enough automated experimental capacity for AI to learn from physical reality as quickly as it learns from text. Adaptyv's goal is for lab throughput to no longer be the constraint on how fast biology can progress.

Benchling has expanded its R&D platform to support chemistry alongside biology. The company launched Reactions, a new capability for planning and modelling single-step synthesis, and added predictive models for small molecule R&D from Inductive Bio. Reactions allows chemists to sketch reactions, automatically calculate stoichiometry, and search past reactions by chemical structure. The platform keeps molecules linked to their synthesis history and experimental results. Benchling's Model Hub now includes Inductive Bio's ADMET prediction models, which help chemists predict chemical and pharmacokinetic properties before physical testing. The models placed first in three consecutive OpenADMET blind prediction challenges. Chemistry capabilities are currently in limited availability. Benchling serves more than 1,300 companies worldwide, including Merck, Moderna, and Sanofi.

Benchling brings chemistry and biology together on one R&D platform, adds chemistry reactions and predictive models. * By Benchling * Aug 20, 2026 Updated Aug 20, 2026 * 0 SAN FRANCISCO, Aug. 20, 2026 /PRNewswire/ - Benchling today announced its expansion into chemistry, bringing chemistry and biology together on one connected platform for scientific research and development. As part of this expansion, Benchling is launching Reactions, a new capability for planning and modeling single-step synthesis, and adding predictive models for small molecule R&D from Inductive Bio. For more than a decade, Benchling has been the market leader at large molecule R&D. As research has evolved, therapeutic modalities are increasingly crossing the boundary between biology and chemistry. In addition, advances in AI create new opportunities to learn from data across the entire R&D organization. This makes a connected, structured data foundation more important than ever. That's why Benchling is bringing molecules and reactions together with experiments, biological results, and computational predictions, so teams work across modalities in one system, and put their scientific data to work with AI. Expanding Benchling for chemistry Chemistry moves in a tight, iterative loop: decide what to make, determine how to make it, synthesize and characterize it, test it, learn, and repeat. But the data behind that loop often lives across disconnected tools, making it hard to understand how a molecule was made and performed, and what to do next. Reactions is an important step in a broader effort to bring more of the chemistry workflow into Benchling, beginning with the foundational workflows chemists use every day. Reactions is built on the same structured platform as the rest of Benchling. Chemists can search across reactions and notebook entries, while every molecule stays linked into the customer's broader scientific record, connecting synthesis with the experiments, results, and other scientific context around it. A chemist designing a new analog can search related molecules and past reactions, plan synthesis with automatically calculated stoichiometry, account for salts and solvates, and capture the resulting product in the same scientific record. When that molecule moves into testing, its history stays connected. Teams trace how it was made, see its experimental and biological results, compare it with related chemistry, and carry those insights into the next design decision. With Reactions, teams can: * Plan synthesis with less manual work. Sketch single-step reactions directly or bring them in from existing reaction files, like ChemDraw, then automatically calculate stoichiometry and quantities as reaction conditions change. * Work with chemically meaningful forms. Use shared salt and solvate libraries and account for salt forms during reaction planning. * Learn from chemistry the team has already done. Find past reactions by chemical structure or similarity, then narrow results by reactants, products, solvents, catalysts, and reaction role. * Work with molecules alongside their synthesis history. Capture and find molecular structures in Benchling alongside the reactions in which they were made or used. * Keep the molecule's story connected. Automatically link reactions to compounds, experiments, and biological results so synthesis history and downstream performance can be explored together. Bringing predictions into the design cycle Benchling Model Hub brings computational predictions into that decision-making loop, alongside the experimental data scientists already use. As we expand our support for small molecule R&D, we're also bringing more chemistry-focused scientific models into Model Hub. Inductive Bio's ADMET prediction models enable any chemist using Benchling to predict the chemical and pharmacokinetic properties that determine whether a molecule can become a viable drug. Inductive Bio's models are independently validated by three consecutive first place performances in the OpenADMET blind prediction challenges. Scientists can register proposed molecules in Benchling, run predictions before physical ADMET testing, and compare those predictions with experimental results from compounds they've already made. That combined evidence helps teams make better decisions about what to synthesize next and move through design cycles faster. We're just getting started Chemistry is currently in limited availability. Request access here: benchling.com/chemistry-la-signup. About Benchling Benchling is the AI platform for biotech R&D, unifying scientific data and automating workflows to accelerate discovery and development. Trusted by more than 1,300 companies worldwide, from pioneering startups to global leaders like Merck, Moderna, and Sanofi, Benchling gives scientists a single place to capture, connect, and act on data across the entire R&D lifecycle. With Benchling AI, agents and models work directly inside scientific workflows, grounded in structured data. The result is faster teams, better molecules, and breakthroughs that reach the world sooner. SOURCE Benchling Submit your news. We're always interested in hearing about news in our community. Let us know what's going on!