Mandrake Bio builds AI-designed gene-editing enzymes for agriculture and therapeutics. The Bengaluru startup has now raised ₹16 crore in a pre-seed round co-led by Activate and Antler, betting that better gene editors won’t come from endlessly tweaking old CRISPR tools but from designing new proteins from scratch. That matters because a lot of today’s gene-editing work still starts with enzymes nature happened to make first, even when those tools are too large, too rigid, or awkward to deliver for crop and medical use. Founded in 2025 by Tanay Lohia and Dr. Kutubuddin Molla, Mandrake Bio is still pre-revenue. It’s already trying to build the kind of AI-plus-wet-lab loop most Indian biotech startups only talk about.
What does Mandrake Bio actually build?
Mandrake Bio is building a full-stack protein-design platform for gene editing. The workflow is pretty specific: it starts with a biological design brief — things like cell type, delivery route, edit precision, and compact size. It then runs those constraints through its internal data layer, called Mandrake Terrabase, and through sequence- and structure-aware models that generate editor candidates. Those candidates are then tested in wet-lab assays. The results feed back into the next design cycle.
That’s a very different pitch from classic CRISPR engineering. Instead of beginning with a natural enzyme and shaving off limitations later, Mandrake wants the desired function built in from day 1. The platform is “function-aware,” which means the model isn’t just throwing out protein sequences at random and hoping a few survive. It’s trained around the biological job the editor needs to do.
The company has also built the surrounding research infrastructure that makes those models useful in practice. Terrabase is its interconnected DNA-protein data system. The wet-lab side validates designed proteins in-house and through partners. That matters because protein design startups live or die on whether computation survives contact with biology. As Lohia put it, “Everybody else starts with an enzyme found in nature and then modifies it. We are building models to design these enzymes from scratch.”
Mandrake also isn’t chasing one narrow use case. Its site points to 2 early application tracks: tissue-culture-free editing of elite crops, which could cut gene-editing timelines by up to 80%, and in vivo gene therapies designed around delivery constraints from the start, with the aim of using lower doses and simpler delivery methods. That’s ambitious. It fits the company’s thesis that compact, programmable editors can outperform inherited CRISPR-era defaults.
Who founded Mandrake Bio and how far along is it?
Founded in 2025 with an AI-meets-biology thesis
Mandrake Bio was founded in 2025 by Tanay Lohia and Dr. Kutubuddin Molla. Lohia is the founder and CEO, while Molla brings the deeper biology expertise as a plant genome engineer and CRISPR scientist who guides the startup’s agricultural applications and wet-lab strategy. That split makes sense. One founder is pushing the computational and company-building side. The other understands how gene-editing tools have to behave in real plant systems, not just in a model.
Lohia’s own bio is unusually blunt for a startup website. He describes himself as a “curious generalist” who fell hard into the rabbit hole of gene editing and proteins, then started Mandrake to build hypercompact editors from scratch. It’s casual language, sure. But it explains the company’s angle. This isn’t a traditional discovery biotech trying to bolt AI onto an old workflow. It’s an AI-native protein design lab trying to rebuild the workflow itself.
Why the founders fit this problem
Molla is probably the clearer signal for market fit. His expertise is in plant genome engineering and CRISPR, which lines up directly with Mandrake’s agriculture-first push. That matters because the first commercial buyers Lohia is targeting are seed companies, not just drug developers. If Mandrake can design smaller editors that work directly in elite crop tissue, that could make plant editing faster and more usable outside research settings.
The broader team also shows what kind of company this is becoming. The roster spans computational biology and structural biology. It also covers CRISPR-Cas systems, metagenomics, AI infrastructure, and wet-lab assay development. That’s the right mix for a startup that needs model quality, biological context, and lab proof all at once.
Where the company stands today
At the time of the round, Mandrake Bio was pre-revenue and had a team of 8 employees. It plans to validate its first set of enzymes in the lab within the next 2 months — which means by September 2026 if it sticks to the stated timeline — before moving into commercial partnerships. Lohia said the first customers should be seed companies and therapeutic companies, with licensing fees forming the basic business model.
Funding, hiring, and the next milestone
Activate and Antler co-led the ₹16 crore pre-seed round, with participation from Spectrum Impact, DeVC, and angel investors Vijay Chandru, Paras Chopra, Sanjiv Rangrass, and Vatsal Dusad. The money will go into expanding Mandrake’s AI protein-design platform. It will also fund hiring across AI and wet-lab research, along with the compute and experimental work needed to validate its proteins. Lohia put it plainly: “A lot of what we build has to be tested in real lab settings.”
The next financing step is already mapped out. Mandrake plans to raise a larger round after it has initial validation data, then use that to set up its own lab and run experiments at greater scale. It’s a sensible sequence. In this kind of biotech, capital gets cheaper once the science stops being hypothetical.
How does Mandrake Bio compare with CRISPR-first rivals?
Most legacy gene-editing companies still begin with natural enzymes — usually CRISPR-derived systems or other naturally occurring nucleases — and engineer from there. Mandrake’s argument is that this approach carries too much biological baggage. Natural editors evolved for microbial survival, not for neat delivery into crop tissue or human therapeutic settings. So the startup is trying to design the editor around the application, not the other way around.
A close global peer is Profluent in the US, which is also building AI-designed gene editors and has pushed that thesis far enough to land a major collaboration with Eli Lilly. That’s useful context because it shows investors are already backing the broader idea that gene editors can be generated computationally rather than merely discovered in nature. Mandrake’s difference is more specific: it’s starting with compact enzymes and agriculture-first use cases. It plans to license into seed and therapeutic companies instead of trying to become a full drug developer itself.
Why are investors backing Mandrake Bio now?
Because this round is really about turning a thesis into data.
Mandrake doesn’t need growth capital yet. It needs proof capital. The money is going into compute infrastructure and wet-lab validation because that’s the only way to show its models can produce enzymes that actually work, not just nice-looking sequences. For early investors, that’s the inflection point worth paying for.
There’s also a practical business reason the round matters. Licensing enzymes to seed and therapeutic companies is a cleaner model than trying to build entire downstream products alone. If Mandrake can validate a first wave of editors by September 2026, it gives potential partners something concrete to evaluate. It also gives the company a much stronger case for the larger round it already expects to raise after the first lab data lands.
How big is the gene editing market Mandrake Bio wants?
It’s not a tiny niche. Grand View Research estimates the global genome editing market could reach $25 billion by 2030, and a newer forecast pegs the broader gene editing market at $39.8 billion by 2033. That kind of growth attracts a lot of startups. It also explains why investors are willing to fund infrastructure bets before revenue shows up. If better editors improve delivery, precision, or cost, they can matter across a huge stack of therapeutic and research workflows.
The agriculture side is big enough on its own to justify Mandrake’s early focus. Grand View pegs the agricultural biotechnology market at $151.23 billion in 2024, with a projected rise to $212.57 billion by 2030. The trend underneath that number is straightforward: crop developers are under pressure to produce plants that are tougher, higher-yielding, and less dependent on pesticide and fertiliser inputs. Gene editing fits that demand. What’s still missing, in many cases, are tools compact and flexible enough to make deployment easier.
Mandrake Bio isn’t being funded because it already won. It’s being funded because the next 2 months could show whether its core claim survives the lab.
Read how Fora Travel raised a $60M Series D at a $1B valuation to expand Via, its AI travel advisor platform, and modernize how independent travel advisors research, book, and manage trips.
FAQ
- What is Mandrake Bio and what does it build?
Mandrake Bio is a Bengaluru biotech startup building programmable gene-editing enzymes with AI and protein design. It uses a workflow that starts with design constraints such as cell type and delivery route, runs those through its internal platform, and then validates candidate proteins in wet-lab assays. - How much funding did Mandrake Bio raise?
Mandrake Bio raised ₹16 crore in a pre-seed round announced on July 16, 2026. Activate and Antler co-led the round, with Spectrum Impact, DeVC, and angels including Vijay Chandru, Paras Chopra, Sanjiv Rangrass, and Vatsal Dusad also joining in. - Who founded Mandrake Bio?
Mandrake Bio was founded in 2025 by Tanay Lohia and Dr. Kutubuddin Molla. Lohia leads the company, while Molla brings plant genome engineering and CRISPR expertise that directly supports Mandrake’s early agriculture push and wet-lab strategy. - Is Mandrake Bio an agriculture biotech company or a therapeutics startup?
It’s both, but agriculture comes first. The company’s enzymes are being built for seed companies and therapeutic companies, though its initial commercial focus is on crop applications such as developing plants that need fewer chemical inputs or handle climate stress better.
