Keep sensitive work where it belongs
Run Pūnaha on customer-controlled infrastructure with explicit model, storage and service placement choices.
Knowledge, AI and work in one place
Find what your organisation knows. Put it to work.
Your organisation already has knowledge, systems and people who know how the work should be done. Pūnaha helps you connect them. Find the right information, use the AI that suits the job, and move the work forward without taking judgement away from the people responsible for it.
Start with your needs
Choose the shape of the system instead of working around the product
You can decide where Pūnaha runs, which models it uses and where a person needs to stay involved.
Use the right AI for each job
Run supported local models or connect external providers. One workflow does not have to depend on one model.
Build around the way the work is done
Connect knowledge, prompts, models, decisions, documents, integrations and people in a workflow you can see.
Let the work change when the situation changes
Use an adaptive case when the next step depends on new evidence or professional judgement.
Leave important decisions with people
Pause for review or approval when context and accountability matter more than speed.
Control that follows the work
Keep the important choices in the hands of your organisation
Pūnaha gives your team practical control over where the system runs, who can use each resource and how work moves between knowledge, AI and people.
You can begin with everything together, or separate parts of the platform when your architecture calls for it. You can also connect the systems and models you already use instead of rebuilding the process around one provider.
Access can be set for a named workflow, case, knowledge store, model or connection. When someone needs to understand a decision, Pūnaha can show why the access was allowed or denied.
Choose how it runs
Use an all-in-one or separated service shape on customer-controlled infrastructure to suit the implementation.
Review deployment choicesChoose who can use it
Give people, groups and services the access they need to named resources, then review that access with clear evidence.
Explore access governanceChoose what it connects to
Use approved connections for business systems, databases, email, local models and external AI.
Explore connectionsFrom knowledge to useful work
A good answer should help someone do something
Finding the right source matters. What happens next matters just as much.
People do not always remember a document title or the exact language used in a policy. Pūnaha can search supported documents, email and connected sources by meaning, exact terms or both.
That source material can then support a person directly or become part of a workflow with visible steps, review points and a record of what happened.
Find
Search trusted organisational knowledge using meaning, keywords or a combination of both.
Explore enterprise AI searchConnect
Choose a supported local or external model for each task and environment.
Explore deployment choicesAct
Turn useful answers into repeatable workflows with human approval when needed.
Explore AI workflowsAdapt
Coordinate case work where people can plan the next appropriate task as new information arrives.
Explore adaptive casesDifferent work needs different structure
Some work follows a route. Some changes as it goes.
Pūnaha does not force a process, a decision and a case into the same kind of diagram.
BPMN-based processes
Model and run structured processes with events, tasks, gateways, waits, subprocesses and durable execution.
Explore BPMN automationDecision tables
Express repeatable business decisions as explicit inputs, ordered rules and inspectable outputs.
CMMN-based cases
Manage evolving work with case files, stages, tasks, milestones, sentries and discretionary planning.
Explore adaptive case managementA simple example
What happens after someone asks a question?
The answer may be simple, or it may need a model, a person and another system. You can see each step.
- 1
Ask
Start with a question, request or event.
- 2
Retrieve
Bring in relevant material from an approved knowledge store.
- 3
Reason
Send the task to the local or external model selected for that step.
- 4
Review
Pause for a person when policy or judgement requires it.
- 5
Continue
Return an answer or pass the approved result to the next step.
Choose your view
Start with the questions you are responsible for
A board member, a technology leader and a developer need different details. Each view leads into the same product.
Board and executive teams
Understand the strategic choices around knowledge, AI, deployment and control.
Read the leadership viewTechnology and operations leaders
Assess deployment fit, governance, models and the path to practical adoption.
Review technical fitDevelopers and configurators
See how knowledge stores, model deployments, workflows, processes and cases come together.
Explore the developer viewProduct tours
See Pūnaha through four recognisable work patterns
These are product examples rather than customer stories, so you can focus on how the system works.
Enterprise knowledge
Find reliable answers across policies, documents and shared knowledge.
Take the tourRegulated research
Gather evidence, use AI assistance and keep a person at the decision point.
Take the tourFlexible AI workflow
Use different models for different parts of the same process.
Take the tourCustomer support
Retrieve approved guidance and prepare a response for review.
Take the tourSee whether it fits
Bring us one problem worth improving
We will shape a private demonstration around the knowledge, process or environment you want to understand.