For biotech and CRO labs
Automated bioinformatics and private AI, installed in your lab.
We install the pipelines and a private AI on hardware you own. Results are ready when a run finishes, and your data never leaves your network.
30 minutes · No cost · No obligation · We'll send a written recommendation afterwards.
Work directly with Filip, who audits, builds and supports your system.
- Best-fit team size
- 5–10 people
- Pipelines and AI run locally
- In your lab
- Illumina, Element, PacBio, Nanopore
- Your instruments
- Checked against your past results
- Validated rollout
Start with one workflow
Take the routine analysis off your team without changing how your science works.
We start with one assay, learn how your team handles it today, and automate that path end to end before adding anything else.
01 · Sequencing run to results
Results and QC waiting when a run finishes.
The system notices a finished run, checks the sample sheet, runs the pinned nf-core pipeline and writes a QC report and standard results to your storage.
Bulk RNA-seq available now. Single-cell, WGS / WES, ATAC-seq and amplicon on request.
- Detect
- Process
- QC
- Report
02 · Questions about your data
Plain-English answers your scientists can check.
Scientists ask about runs, samples and genes. The AI reads the pipeline outputs and your documents, and every number links back to its file and row.
The pipeline does the statistics. The AI explains results; it never computes them.
- Ask
- Look up
- Cite
- Answer
03 · CRO sponsor projects
Each sponsor's data kept apart, every run on record.
Per-project access, a run record for every result (inputs, pipeline version, parameters) and reports you can hand to a sponsor.
Access is enforced by the system, not by instructions to the AI.
- Isolate
- Process
- Record
- Deliver
Your scientists keep the scientific judgment. The system prepares results and answers; people decide what they mean.
What it does for your lab
Your analysis becomes lab infrastructure, not another metered cloud bill.
You buy the hardware. We implement and maintain it. The pipelines are open source, the models run under their own licences, and your data, results and configurations stay yours.
Compare private and cloud AI- 01
Results the day a run finishes
Processing, QC and standard reports start on their own. No queue, no waiting on one person.
- 02
Time back for your scientists
Routine processing and quick questions stop landing on your bioinformatician, or on nobody.
- 03
Confidential data you can use with AI
Sequencing data, results and questions stay on your network. Nothing goes to an outside AI service.
- 04
Results you can trace and re-run
Every result is recorded with its inputs, pipeline version and parameters.
- 05
A cost you can plan
You own the hardware. Support is one flat monthly fee, with no per-sample, per-user or per-question charges.
- 06
Scope that stays clear
New assays and major additions are quoted separately, never hidden inside an unlimited-support promise.
Use what already works
Connect to the instruments and systems your lab already relies on.
We start from your current setup and connect only what the first workflow needs.
Your systems
- Sequencer run folders
- Lab storage (NAS)
- Sample sheets
- LIMS / ELN exports
- Papers and SOPs
Approved · Read access + company login
Cerebra node
Pipelines + private AI
Runs in your lab
Recorded · Run records + audit trail
What your team gets
- QC reports
- Expression and DE results
- Cited answers
- Run records
- Failure alerts
Connections and permissions are agreed workflow by workflow. Your existing storage stays the system of record.
What we build
Everything needed to run it in your lab, set up and supported.
We specify the hardware you buy, then install, wire and tune the whole stack around your assays and your team.
- 01
Hardware
A compact AI computer and network storage, sized from your run volumes. Bought by you at vendor price.
- 02
Data and network
Connection to your sequencer output, storage layout, backups and your company sign-in.
- 03
Pipelines
nf-core pipelines for your assays, versions pinned, started automatically when a run finishes.
- 04
Local AI
An open model served on your hardware, tuned for speed, connected read-only to your results and documents.
- 05
Access and records
Per-project access, an audit trail and a record of every run.
- 06
Your team
Training for scientists and your admin, a runbook and documentation.
How pricing works
- Audit
- Free
- Pilot on one assay
- Fixed fee
- Setup and installation
- Fixed project fee
- Support and maintenance
- Flat monthly fee
- Hardware
- Bought by you, no markup
Quoted after the audit. No per-sample or per-user charges.
What each fee includesThe process
From first call to a system your team uses every day.
Seven steps. We understand the work first, then automate only what passes checks against your own results.
Step 1: Pick the first workflow
Choose the assay where automation saves the most time, and agree what is out of scope.
Step 2: Gather what exists
Wet-lab protocols and library prep kits, run folders, sample sheets, current scripts, past results, SOPs and input from the people doing the work.
Step 3: Map the lab and the analysis
Every step from sample to report: how samples are prepared and sequenced, then the tools, parameters, QC thresholds and where outputs go.
Step 4: Set controls and success criteria
Agree access, sign-in, what the system may read and write, and how results will be judged.
Step 5: Build and connect
Install the hardware, pin the pipeline versions and connect only the storage and systems needed.
Step 6: Validate on your data
Re-run recent runs and compare them with your existing results before anyone relies on the output.
Step 7: Operate and improve
Train your team, monitor the system and add assays once the first workflow is stable.
Ongoing support
The system isn't handed over and forgotten.
Filip stays directly responsible for keeping pipelines, models and hardware working as your data and team grow.
Flat monthly support
- Monitoring
- Hardware, storage, pipeline and AI model health.
- Maintenance
- Security updates and fixes, in maintenance windows you agree to.
- Pipeline and model updates
- Tested against your past results and known questions before they replace the current version.
- Performance tuning
- Keep runs and AI responses fast as data volume and users grow.
- Direct support
- Work with the person who built the system.
- Training
- Help scientists and admins use the system, and other approved AI tools, well.
New assays and material expansions are scoped separately.
Security
Designed to pass your IT review.
Stays on your network.
Data, results and AI conversations never leave your building.
Your company login.
Sign in with Microsoft or Google accounts your IT team already manages.
You control our access.
Remote support is off by default. Your admin switches it on when needed, and every session is logged.
Full audit trail.
Every login, question, pipeline run and admin action is recorded.
Filip Isakovic
Founder, Cerebra BioSystems
About Filip
Built and supported by the person whose name is on the page.
Filip Isakovic founded Cerebra BioSystems and does the work end to end: auditing the lab, designing and installing the system, supporting the rollout and staying your direct point of contact.
Filip is a bioinformatician who has built analysis pipelines and LIMS and ELN systems for lab teams. The focus is practical, private analysis systems for teams that handle confidential data and can't treat it as a generic software subscription.
More about FilipFAQ
Common questions.
Does any data leave our building?
What do we own?
Do we have to replace our current software or storage?
How much does the system do on its own?
What does the monthly fee include?
What size of team is the best fit?
Bring the analysis your team is tired of doing by hand.
In a free 30-minute audit we'll map your instruments, data volumes and analysis needs, then send you a written recommendation, whether or not you work with us.
30 minutes · No cost · No obligation · We'll send a written recommendation afterwards.