The thinking system for operational decisions
ThinkingOps turns your SOPs and institutional knowledge into AI agents that work every case against your rules. Our engineers build them on Blackett, our platform, and run them with your team, so nothing goes unchecked and every decision stays yours.
Scaling your operations without scaling the busy work.
Your best experts spend the day stitching data together
They copy figures between screens and checkeach one against the SOP, the policy and theregulation. It is the most repetitive part of the job,and it takes most of the day.
The data is spread across systems
Records, calls, contracts and spreadsheets each hold part of the workflow. When nobody sees all of it, two people reach two different answers on the same data.
Manual checks do not scale with the work
As volumes grow, your team can check only a sample. Checking everything by hand would take a person on every check.
Blackett, the operating system that powers every workflow we build for you
Every workflow runs on the same operating system, so your second one starts where your first left off, with your rules, tools and data already connected.
We build the workflow for you
We do the work for you
Query the work. See its rationale. Correct the work and give feedback in the same interface.
Operations you can rely on. Outcomes you can prove.
Transform complex workflows into learning operational systems.
In 1mg’s quality audits, doctors and the quality teamreview every case, with faster turnaround.
Blackett’s context layer with open-weight models, compared with frontier models alone.
On healthcare digital quality measures, checked against known answers.
Start from a workflow like these, or bring us your own
We build each workflow with your team, on Blackett, fitted to your data, rules and tools. Start with the busy work that takes your experts’ day, and let them get back to the work only they can do.
Life sciences operations
Checked against the measure spec, the payerpolicy and your own SOPs.
Every chart, note, call and claim
Checked against the measure spec, the payer policy and your own SOPs.
Every chart, note, call and claim
Checked against the measure spec, the payer policy and your own SOPs.
Every chart, note, call and claim
Checked against the measure spec, the payer policy and your own SOPs.
Quality measures (HEDIS, MIPS)
Every chart is checked against the measurespec, and each answer is linked back to theline in the record that supports it.
Quality measures (HEDIS, MIPS)
Every chart is checked against the measurespec, and each answer is linked back to theline in the record that supports it.
Quality measures (HEDIS, MIPS)
Every chart is checked against the measurespec, and each answer is linked back to theline in the record that supports it.
Build your own
Bring us an operational workflow. Wemap it with your team and build it onBlackett.
What customers say
ThinkingOps’ AI-driven Quality audits, letour doctors and quality team review 100%of cases with smoother operations and faster turnaround.
ThinkingOps transformed our site enrollment workflow - from protocol
ingestion to cohort qualification - into one AI-powered experience, helping us move faster with fewer screen failures.
ThinkingOps helped us make cancer detection more accurate, scalable, and affordable with Highly accurate AI & ML Models.
Builders and operators who have run this work themselves
Our operators have run operations at scale in payments, healthcare delivery and enterprise software, from calls to transactions and from notes to claims.
Our AI engineers have built data and AI products at Palantir and Microsoft. Together they build every workflow with you.
Core Team
TalentSprint. Built and scaled
Blackett’s context layer.
.svg)
Your data, your rules, your sign-off
Privacy and security
SOC 2 Type II audited, HIPAA and GDPR compliant, with end-to-end encryption on every record we touch
Runs in your environment
Self-hosted in your VPC or private cloud, so your data never leaves your environment
Audit-ready
Source, rule version, reviewer and outcome kept for every workflow run
People approve what you choose
Decide which work closes on its own and which waits for a person
Works with what you run today
Connects to the systems you already use and the AI models you approve, and sends results back where your team works.
Frequently Asked Questions
Does our data leave our environment?
No. Blackett runs in your VPC or private cloud, and your records stay there. SOC 2 Type II, HIPAA and GDPR, with complete audit trails.
What happens when the AI gets it wrong?
You choose which work waits for a person. Sources, rules and rationale make outputs reviewable. Your team can correct the work, and that feedback improves Context and Reasoning. Workflows are tested before go-live; human review and testing do not mean errors are impossible.
Do the agents act on their own?
Only where you allow it. You set the mandate for each workflow, including which actions can run unattended and which need approval. If you require human review, that boundary remains in place as the system learns. Work needing approval routes to your team with its context attached.
Is this software or a service?
The rule is changed once, in Blackett’s rulebook, and every workflow that uses it follows the new version. The audit trail records which version was used for every workflow run. There is no rebuild and no re-implementation cycle.
How is this different from building it ourselves on an AI model?
The model is the easy part. Blackett gives every workflow the rules, agents, evidence trail and evaluation that a custom build has to create from scratch, and our operators already know the edge cases that break a rule.
Which industries do you work in?
Examples include healthcare operations, finance operations, revenue operations and life sciences. The common factor is repetitive operational work across data and written rules ,not one industry. Bring your own workflow for us to map with your team.
Where do we start?
With one operational workflow. We map the data, rules and mandate with your experts, build it on Blackett and tune it with your team before go-live.
