swivl Let’s talk

MEET TES · BY SWIVL

AI that carries
responsibility.

Work moves on. Intelligence should stay.

Meet Tes. Persistent organisational intelligence that keeps knowledge current, acts within human-set limits, and learns from verified outcomes.

Built in Finland. For the life of your business.

ONE RESPONSIBILITY. AS THE WORLD CHANGES.

Keep the network moving.

An EV charging operator’s continuing responsibility.
Illustrative experience · not a live product demo

01 / THE COMMITMENT

“Restore service.
Keep me informed.”

The outcome stays with Tes, along with the people, history and authority needed to carry it forward.

Responsibility
Restore charging at the affected site
Authority
Repairs up to €1,000
Success means
Charging works again—not just a closed ticket

02 / THE FACTS CHANGE

The repair is bigger
than expected.

A technician’s diagnosis changes the plan. Tes connects the new evidence to the existing commitment.

Earlier assumption
Repair within the €1,000 limit
New evidence
Replacement required · quote €1,400
Consequence
The current authority is not enough

03 / THE PLAN ADAPTS

Keep the work moving.
Keep the boundary.

Tes explains the changed cost, requests approval and coordinates permitted next steps. She doesn’t quietly raise her own limit.

Continue
Coordinate availability and update the operator
Ask the principal
Approve the €1,400 replacement?
Unchanged
The spending limit remains €1,000

04 / THE OUTCOME IS CHECKED

“Done” needs
more than a message.

After approval and repair, Tes checks the result against the original commitment before treating the work as complete.

Authority recorded
Operator approved the replacement
Work reported
Technician installed the part
Outcome verified
Successful charging session confirmed

05 / EXPERIENCE STAYS

The next response
starts further ahead.

Tes retains what worked, why it worked and when it applies. A changed context calls for a fresh check—not blind repetition.

Retain
Diagnosis, repair and verified result
Conditions for reuse
Matching fault, equipment and operating context
Still human-set
Learning never grants new spending authority

THE TECHNOLOGY

Models power the reasoning.
Tes carries the continuity.

Our architecture puts the organisational decision loop in Tes: what matters, what is authorised, what changed, and what to learn.

  1. 01State

    What is true now?

  2. 02Decide

    What matters next?

  3. 03Act

    What is authorised?

  4. 04Verify

    What actually changed?

  5. 05Learn

    What deserves reuse?

Verified outcomes revise the state of the work
01 / EVIDENCE-AWARE MEMORY

Revise knowledge.
Don’t just accumulate it.

New evidence updates organisational state while preserving the reasons behind decisions.

New evidenceRevised understanding
02 / GOVERNED ACTION

Earn autonomy.
Respect human-set limits.

Mandates, permissions and demonstrated performance govern what Tes and her specialist agents may do.

Mandate + evidenceAuthorised action
03 / OUTCOME-ATTRIBUTED LEARNING

Learn from what worked.
And know why.

Connect verified outcomes to the procedures actually used, making experience available for reuse.

Verified outcomeReusable procedure
04 / CONTRACT-AWARE ADAPTATION

Keep the experience.
Recheck the assumptions.

When operating conditions change, reassess learned procedures before applying yesterday’s answer.

Changed conditionsRevalidation
Organisational capability should outlast a model.Frontier reasoning and tools underneath. Persistent knowledge, authority and learning in Tes.

ONE ARCHITECTURE. MANY RESPONSIBILITIES.

Start with work
that needs continuity.

Not another system to keep updated. Intelligence built around the responsibilities your team already carries.

OPERATIONS

Keep a service running.

Connect incident history, service commitments and authorised responses. Carry the work through to a verified resolution.

DELIVERY

Keep a promise to a customer.

Retain the decisions, dependencies and owners behind a delivery. Reconcile changes without starting the briefing again.

COMPANY BUILDING

Turn an idea into an operating product.

Carry requirements across research, development and evaluation. Coordinate specialist agents while keeping the outcome in view.

Illustrative applications of the Tes architecture.

RESEARCH & EVIDENCE

A technical foundation.
A testable ambition.

Our research spans memory, governance and learning. We bring those mechanisms into working prototypes, then test them against the demands of continuing responsibility.

DEMONSTRATED IN PROTOTYPE

Continuity across restart.

Retained conversation state supports continuation after a restart.

The question: can the work survive a session?

DEMONSTRATED IN PROTOTYPE

Tes directing specialist agents.

Tes delegates bounded work and brings the agents’ results back into the conversation.

The question: can Tes coordinate, not just answer?

NEXT VALIDATION FRONTIER

Responsibility over time.

Measure continuity, supervision, cost and learning as real organisational work changes.

The question: does capability improve with experience?

From individual mechanisms to sustained, real-world responsibility.

Explore a design partnership

THE PEOPLE BEHIND SWIVL

We’ve lived the work
we want to change.

Scientific research, international company-building and strategic finance. A founding team shaped by the reality of growing organisations.

Niina Suuriniemi Hopper

Niina Suuriniemi Hopper

CEO & co-founder

Cambridge PhD and medical-research background. Co-founded EVA Global. Leads Swivl’s customer strategy, commercial validation and investor readiness.

Meet Niina
Donald Hopper

Donald Hopper

Co-founder · Product & technology

Co-founded and led EVA Global through international growth. Leads Tes’s product direction, decision-loop architecture and technical delivery.

Meet Donald
Brian Erola

Brian Erola

Co-founder · Strategic finance

Former finance leader at EVA Global, Digital Charging Solutions and BMW Group. Brings forecasting, commercial pricing and business modelling to Swivl.

Meet Brian

LET’S BUILD WHAT COMES NEXT

What would you do with
intelligence that stays?

Talk to Niina

niina@swivl-ai.com