The company that decided a lab test is really a search engine - and rewrote the math underneath it.
The logo does the thesis in one image: a double helix stitched out of ones and zeros. They mean it literally - biology, read as data.
Molecular tests are brilliant at finding a needle in a haystack. Algorithmic Biologics asked the more annoying question: what do you do when you have a thousand haystacks and only a few of them have needles?
Here is a fact about molecular testing that sounds like a paradox but is really just arithmetic. If you have 1,000 samples and you suspect only a handful carry the thing you are looking for, you do not actually need 1,000 tests. You need enough tests to locate the handful. The rest is redundant - you are paying to confirm that mostly nothing is there.
This is the observation that Algorithmic Biologics, a deep-tech company with one foot in Bengaluru and one in the United States, has turned into a business. Its founder, Dr. Manoj Gopalkrishnan, is not a biologist. He is a computer scientist - a B.Tech from IIT Kharagpur, a PhD from USC, and teaching stints at USC, Duke, and IIT Bombay - who spent roughly eighteen years on a field called molecular computing. The premise of that field is that chemistry can process information, the way a computer does. Most people hear that and nod politely. He built a company on it.
The company's core product is called Tapestry, and the elevator version is this: it lets a lab pool many samples into a few pools, run those pools on the machines it already owns, and then use math to figure out which original samples were positive. The math is not hand-waving. It borrows from compressed sensing - the same principle that lets a camera or a phone reconstruct a full signal from far fewer measurements than you would naively think you need - and from combinatorial group testing, a wartime idea for screening many blood samples at once.
What makes it more than a clever trick is that naive pooling usually loses information. Mix ten samples, get one murky answer, and you often cannot tell which sample was positive or how strong the signal was. Tapestry's pooling patterns are designed so the answer stays recoverable - quantitatively, in a single round of testing. That distinction, boring as it sounds, is the whole company. It is the difference between an idea that demos well and one that a regulator will approve.
And regulators did. Tapestry first proved itself during COVID-19, when the whole world suddenly needed to test enormous numbers of people cheaply. It became DCGI-approved and CE-certified, was validated on more than 15,000 samples at Indian and international labs, and was the only indigenous Indian technology to reach the finals of the XPRIZE for Rapid COVID Testing - a competition otherwise dominated by hardware. Software, it turned out, could compete with instruments by simply making the instruments smarter.
"Bringing ideas of AI and molecular computing to testing could enable better health for all."
Software tells the lab which samples to mix into which pools - following a designed pattern, not a random one. Far fewer tubes than samples.
The pools run on the lab's existing assays and instruments - qPCR, NGS, mass spec. No new hardware, no new protocol to certify.
The cloud algorithm deconvolutes the pooled signals - with noise modeling - back into per-sample, quantitative results.
The savings grow as positives get rarer - which is exactly the regime that screening lives in.
Cloud-delivered, single-round quantitative pooling. Test many samples in a few reactions using compressed sensing plus combinatorial group testing - on assays you already run.
A compression architecture for multiplexed assay design. The company cites 5x more assays launched at ~10x lower cost - shrinking development from roughly 18 months to 3.
Applied to NGS workflows to remove an estimated 60-90% of library-preparation cost - the hidden tax that makes transcriptomics expensive before you learn anything.
In plain terms: if you run a diagnostics lab, a sequencing service, or a pharma screening pipeline, the pitch is that you keep your instruments and your protocols, add a layer of math on top, and get more answers per dollar. The technology reaches beyond healthcare too - the platform has cited uses in agriculture, food safety, animal husbandry, and synthetic biology research.
Molecular tests are great at finding a needle in a haystack. Tapestry is great when there are many haystacks and few of them have needles.
- Dr. Manoj Gopalkrishnan, Founder & CEO| Legal name | Algorithmic Biologics Pvt Ltd |
| Founded | 2021 |
| Founder / CEO | Dr. Manoj Gopalkrishnan |
| HQ | Bengaluru, India + US office (Delaware / California) |
| Team | ~17 people |
| Total funding | ~$2.84M (Seed) |
| Backers | Bharat Innovation Fund, Axilor Ventures, BIRAC |
| Sector | Molecular diagnostics / deep tech |
The founder is a computer scientist, not a biologist. The whole company rests on the idea that a test tube is a kind of processor.
The core math - compressed sensing - is the same principle that lets your phone reconstruct an image from very few samples.
Naive COVID pooling failed in many places because it loses information. Getting the pooling pattern right is the entire moat.
Seventeen people, eight patents, a $60B addressable market. A small team betting on one principle instead of one product.
Algorithmic Biologics is a Bengaluru- and Delaware-based deep-tech company that treats molecular testing as an information-processing problem. Founded by molecular-computing researcher Dr. Manoj Gopalkrishnan, its patented Tapestry platform uses compressed sensing and combinatorial pooling to run large-scale DNA/RNA/protein tests with far fewer reactions - screening many samples at once and only zooming in where a signal exists. The approach first proved itself as an award-winning, regulator-approved COVID-19 pooling solution and now targets affordable genomics, multiplexed diagnostics, and accelerated assay design.
Last updated: