Bioqore: AI-Guided Process Development for Biologics, Proven in Its Own Lab
Bioqore pairs Bayesian models with an automated wet lab to find process conditions that hold up at scale. Both of its services are on the Cypher AI Marketplace.
By Cypher AI ·
A biologic that works in discovery still has to be manufactured reliably at scale. Teams test cell lines, media and bioreactor conditions to develop that process, but conditions that work at one scale may not perform the same way at the next.
WHAT IS BIOQORE?
Bioqore is a biomanufacturing company that pairs Bayesian AI models with its own automated wet lab to find process conditions for biologics. As of October 2026, two Bioqore services are listed on the Cypher AI Marketplace, a catalog of lab services that scientists can browse and order from.
Josh Hinckley, PhD, is Bioqore's CEO and co-founder. He works in computational biology and bioprocess development, with a PhD from Cornell and postdoctoral work at MIT and the Broad Institute. Janette Phi, co-founder and chief business officer, brings more than 30 years of experience commercializing life-science platforms, including senior roles at Deepcell and ForteBio.
WHY DOES A PROCESS MODEL NEED A WET LAB?
A process model needs a wet lab because its predictions stay hypotheses until a run confirms them. That matters most at scale-up, which is usually where bioprocess plans fall apart.
Bioqore runs that check as a repeating cycle. Its models start from a team's earlier data to map which conditions are worth exploring. They then pick the most promising untested conditions and predict how each will perform, with a 95% confidence interval. Bioqore's automated lab runs those conditions, and the results go into the next round of predictions.
The method has a published record. In a 2025 Nature Communications paper, Hinckley and colleagues at MIT and the Broad Institute used Bayesian optimization to develop cell culture media for human PBMCs and for recombinant protein production in the yeast Komagataella phaffii. They found conditions that beat the standard media with 3 to 30 times fewer experiments than a standard design of experiments would need.
The savings grew with the number of factors in play, reaching 10 to 30 times with nine factors. This can be huge for small teams, because fewer runs per decision can mean months saved on the way to an IND.
WHAT CAN YOU ORDER FROM BIOQORE ON CYPHER AI?
As of October 2026, Bioqore lists two services on the Cypher AI Marketplace. Both are priced by quote, with an estimated turnaround of 1 to 3 months.
AI-Guided Biologics Process Development starts from a DNA or protein sequence, a stalled biologic asset or a sparse process history. Bioqore's Voyager platform turns that starting point into a development plan covering construct design, upstream bioreactor conditions, downstream purification and analytical characterization.
Scale-Up, Validation, and CMC-Ready Tech Transfer takes a process from microscale through pilot scale. It ends with a package the client owns: process recipes, experimental datasets, predictive models, characterization results and CMC-ready documentation. A team can take that package to a CDMO, run the process in-house or use it in a regulatory submission.
HOW DOES A BIOQORE PROGRAM MOVE FROM BENCH TO BIOREACTOR?
A Bioqore program runs in stages, and each stage is a two- to four-week cycle that ends in a report and a go/no-go decision. The stages start with high-throughput media screens and 100 µL to 3 mL microbioreactors. They move through 50 to 250 mL flasks and 1 L bioreactors, then 2, 10 and 50 L runs, and end with the data package and strain handoff.
Each stage adds data to the model before the team commits to a larger run. If a clone performs differently from the model's prediction at an intermediate scale, the team can investigate before moving to the next stage.
Cell line development shows what that can save. In a program with Leadgene Biosolutions, Bioqore took CHO cell line development from single cells to 1 L bioreactor readiness in three months. Bioqore compares that with 9 to 16 months or more for a comparable CDMO engagement.
FREQUENTLY ASKED QUESTIONS
Who owns the data from a Bioqore program? The client does. Clone recommendations, the full experimental datasets and the process data package all go to the client, with no data held back.
Does Bioqore work with microbial systems as well as mammalian cells? Yes. Bioqore's staged framework is active in CHO, HEK and Pichia pastoris, and in human immune cells including PBMCs, CAR-T and NK cells. Its work also covers enzyme strain development and fermentation scale-up for alternative proteins.
What does Bioqore need to start a program? Bioqore asks for the drug candidate, shipped under a material transfer agreement (MTA). It also asks for the team's manufacturing goals, operating limits and any earlier experimental data.
HOW DO YOU ORDER FROM BIOQORE ON CYPHER AI?
You can browse Bioqore's listings at cypherbio.ai/services without an account. To request a quote, create a free Cypher AI account. The request form for scale-up work asks for a project overview, covering goals, molecule type and development stage, and how many strains, proteins or cell lines the work involves.
Bioqore scopes the program and sends you the quote. You pay Bioqore directly.