At Kinatico, artificial intelligence did not arrive as a shiny new technology project
looking for a problem to solve. It arrived as a people project with a practical
mission: give skilled teams more time for the work that really needs human
judgement.
That philosophy has now produced Vera, Kinatico’s first internal workflow AI agent,
built to handle high-volume, rules-based tasks around the clock. Vera verifies
work, provides recommendations, assigns confidence scores, leaves a full audit
trail and escalates to a person when judgement is required. In short, she does the
repetitive heavy lifting without pretending to replace the people who know the
work best.
From Efficiency Track to Agent Factory
Kinatico’s AI strategy runs on two tracks: today’s business made better, and
tomorrow’s business made possible. Vera began squarely on the efficiency track,
where the team had identified repeatable, transactional work that was
consuming time and expertise.
“We wanted capacity that grows with demand without asking more of our
people,” said Odelia Sarre, Kinatico’s Chief Operating and AI Officer.
“Vera works 24/7, so work that used to wait for business hours now moves overnight. She helps us absorb peaks without adding hours to the team, and she gives time back for the higher-judgement work that needs a human.”
Vera’s origin story is refreshingly human. She started as a side project led by three
women: Odelia, a female engineer and a female domain expert, who worked step
by step to turn a real operational need into a production-ready AI agent. The
name came naturally because Vera’s early work focused on verifying. Later, the
team discovered the name’s connection to Veer, a Hindi and Sanskrit word
meaning brave or courageous. It stuck.
Built Fast, Built Responsibly
The result came quickly. Vera was in production and delivering benefits within six
weeks, and she has since become the foundation for Kinatico’s broader “Agent
Factory”. The speed matters, but Odelia is clear that speed was only possible
because governance, safety and trust were designed in from the start.
Trust Was the Real Technology Challenge
For Odelia, the hard part was never simply choosing a model or connecting
systems. The real test was whether people would trust the agent, understand her
limits and feel safe enough to question her. “AI only works when people accept it,
and acceptance really means confidence,” she said. “People need to be confident
it is effective, efficient, ethical and safe. You cannot build that confidence without
psychological safety first.”
Kinatico brought sceptics, supporters and people who felt exposed by AI into the
process early. Legal and privacy teams were involved from day one, risk was
treated as a design input, and Vera went through independent security testing
before live work. Governance, in this story, was not a handbrake. It was the reason
the builders could move quickly without running off the road.
The Efficiency Payoff
The productivity story is easy to understand: Vera does not need office hours,
morning coffee or a quiet spell in the inbox. She keeps work moving overnight and
through peaks, acting as a consistent second set of eyes that can catch near
misses before they become errors. “She is efficient because she is focused,”
Odelia said. “Vera handles the high-volume, rules-based steps, while our people
stay accountable and concentrate on the work that needs experience, empathy
and judgement.”
A Colleague, Not Just a Tool
Staff reaction has been one of the more telling parts of the experiment. People
talk about Vera as a reliable colleague, not just another platform, because the
people who do the work helped shape what she does and how she behaves. Her
weekend work ethic has also won fans.
Just as importantly, Vera has helped lift the conversation about AI across the
organisation. Kinatico measures its AI culture across culture and adoption, vision
and governance, enablement and capability, and impact and value creation.
Across the organisation, all four categories score 80 or more. Overall employee
engagement has also risen by 17 points, taking Kinatico above global
benchmarks.
Lessons for Other Organisations
For organisations considering their own agent, Odelia’s advice is surprisingly low-
drama: start small, start with the right people and start with repeatable work.
Redesign the work before automating it. Bring in the sceptics as well as the
champions. Keep a named person accountable. And keep asking both questions:
could we, and should we?
“Responsible AI is not the slow, cautious option,” Odelia said. “It is how you move
faster with confidence. Vera proved that when you put people, governance and
practical value together, you can build an agent that is not only useful, but
trusted.”
For Kinatico, Vera is no longer just a brave first experiment. She is proof that the
future of work can be efficient without being impersonal, automated without
being unaccountable and intelligent without losing its human centre.