How Kalei transformed Babel's fragmented AI usage into a proprietary, governed AI system

Babel's people were already using Artificial Intelligence (AI). But it was fragmented, disjointed and siloed. It was a chatbot not an organisational context layer. Kalei's work was to connect Babel's AI into one structure: anchored to business value, elevated through organisational context and designed to unlock human creativity. AI provided the ingredients. Babel's people still cooked the dish.

BabelPowered byKALEI
BabelAI HubPowered byKALEI
HomeStrategyAgentsWorkflowsAI ToolsOpportunitiesKnowledgeGovernance
Babel × Kalei
The Babel AI Hub
One front door. Everyone empowered.
Workflows
Agents
Knowledge
Governance
Illustrative view · client detail withheld
Built around your people, not the technology.
kaleilabs.com/case-studies/babel

What were Babel's core operational challenges before partnering with Kalei

Babel, a specialist B2B technology PR and marketing agency, faced severe AI fragmentation. While 75% of Babel's team used AI daily, the use lacked shared standards, centralised memory, and governance. Kalei partnered with Babel to transform these isolated, prompt based workflows into a unified, proprietary AI system.

Babel's industry reputation is driven by the agency's (and its people's) ability to translate technical brands and information into memorable, attention grabbing campaigns and headlines. By the time Kalei began the engagement, nearly three-quarters of Babel's team were already using AI every day.

But the AI use sat in individual workflows: different methods, no shared standard, no shared context, no shared memory and no common best practice. This AI use consisted of a number of 'power users' experimenting with AI's potential and sharing fragmented prompts with colleagues.

This scenario represents what early success with AI usually looks like. The momentum outruns the system around i, resulting in no governance, no shared model and no best practice. Just an ever expanding list of point-solution prompts saved down in a word doc (at best). Kalei identified the opportunity to connect these disparate ideas into a governed system that Babel could own and make unique to the Babel business.

The opportunity for Babel was not just to add more AI. There was already plenty of ideas and prompts within the business. It was to connect it to the right pieces into an ownable, governed AI system, setting the team up for success in the evolving AI-enabled workplace.

Before · individual workflowsNo shared standard
prompts.docx
power user
personal notes
shared prompt
power user
prompt v7 final
method — ad hoc
prompt list
Momentum, without the system around it

How Kalei implemented a 5-step AI strategy for Babel

Kalei deployed a structured, 5-step methodology to build Babel's AI architecture. This process prioritised high-impact business value, preserved human expertise, embedded Babel's proprietary knowledge layer, maximized Babel's existing technology stack, and integrated strict data governance by design.

When deploying AI, its easy to rush into the tools and technology solutions available. But this can be a costly mistake and can see opportunities missed and an AI value layer looking more like spaghetti junction than a cohesive strategy. To prevent a disjointed AI strategy, Kalei executed a rigorous, step-by-step method tailored for Babel:

01

1. Defining the value and strategy

Working with Babel's senior leadership team, Kalei helped to set the business value at stake, what matters most and where the biggest impact would be.

Key to this was defining the defensive (efficiencies) and offensive (innovations) implementations of AI. You can't do everything all at once, so creating a priority list of improvements was essential. Babel actively chose to implement AI workflows that targeted high impact on efficiencies and set the foundations for future innovations and market opportunities.

BabelAI HubIllustrative viewKALEI
HomeStrategyAgentsWorkflowsAI ToolsOpportunitiesKnowledgeGovernance
Value at stakeAgreed upfront
What matters most
Where the biggest impact would be
Defensive · EfficienciesPriority
01
In build
02
In build
03
Scoped
Offensive · InnovationsFoundations
01
Next
02
Later
03
Later
The value at stake, agreed before a single tool — detail withheld
02

2. Human powered, AI enabled design

Kalei determined the exact human involvement required in Babel's workflows. Because Babel is an expertise-heavy knowledge organisation, human judgment had to be preserved by design, as the human layer within the workflow is where the magic would happen and the value created. Kalei mapped how Babel's team thinks and decides. The resulting principle: AI provides ingredients like research and planning, while the human owns the dish - final output. Outputs remain human. AI enhances the inputs.

Ingredients · provided by AI
ResearchPlanning
The person
owns the dish
Human judgement
Outputs remain human
AI enhances the inputs · human judgement preserved by design
03

3. Embedding Babel's contextual knowledge

Through interviews and knowledge-base analysis, the Kalei team built a structured context layer that surfaces Babel's organisational know-how. The know-how that used to live in people's heads. Kalei embedded years of Babel's industry best practice directly into AI-enabled workflows and standalone agents. Consequently, every AI output draws from the same governed, privacy-first context without requiring repetitive background prompting.

BabelAI HubIllustrative viewKALEI
HomeStrategyAgentsWorkflowsAI ToolsOpportunitiesKnowledgeGovernance
Shared context
Embedded in workflows
Best practice
Embedded in agents
Company know-how
Shared memory
Governed, privacy-first context · no repetitive background prompting
One context layer, drawn on by every output — detail withheld
04

4. Maximising the pre-existing tech stack

Technology came last. The parameters made it simple: maximise the tech stack Babel already possessed. Kalei conducted a rigorous analysis to identify the right tool for the defined job - agents, knowledge layer, research enablers, etc - not a platform looking for a problem. Kalei built the architecture to be technology-agnostic and IP model-portable. This enabled Babel to codify and own its proprietary workflows, processes, know-how and context, rather than it be locked inside of any one vendor's model. Meaning it can be applied on top of whichever AI capabilities fit best as the market moves, its tech stack evolves or new solutions purchased.

BabelAI HubIllustrative viewKALEI
HomeStrategyAgentsWorkflowsAI ToolsOpportunitiesKnowledgeGovernance
The right tool for the defined jobExisting stack
AgentsExisting provider
Knowledge layerExisting provider
Research enablersExisting provider
Technology-agnosticIP model-portableNo vendor lock-in
Tools mapped to jobs, on the stack Babel already owns — detail withheld
05

5. Governance built-in, not on top

With each step, governance was layered in, defining clear ownership, decision rights, policies, data handling and privacy rules that travel with the work. Baked in from the start to ensure client-confidential material and proprietary knowledge stays protected at every step.

BabelAI HubIllustrative viewKALEI
HomeStrategyAgentsWorkflowsAI ToolsOpportunitiesKnowledgeGovernance
Travels with the workBuilt in
OwnershipNamed owner
Decision rightsNamed owner
PoliciesNamed owner
Data handlingNamed owner
Privacy rulesNamed owner
Protected at every step — detail withheld

What New Market Opportunities Did Kalei Unlock for Babel?

Beyond driving internal efficiencies, Kalei facilitated an innovation roadmap that unlocked a tiered pipeline of new market opportunities for Babel. This roadmap shifted Babel's focus from doing work faster to offering entirely new, AI-enabled PR and marketing services.

With one eye to the future and what AI could enable, an innovation session Kalei ran with Babel's leadership shifted the question from how do we do this faster and better to what can we offer now that we couldn't before.

To stay competitive in the PR and marketing industry, new services, value offerings and opportunities must continue to stay at the forefront of an agency owners mind. As a traditional 'time and resources' model, the industry is set for significant changes as AI is rolled out.

Working with the Babel team, they now have a tiered pipeline of new market opportunities. Ranked and prioritised based on the biggest impact for the investment, giving them a structured roadmap to move the company forward as AI takes centre stage.

BabelAI HubIllustrative viewKALEI
HomeStrategyAgentsWorkflowsAI ToolsOpportunitiesKnowledgeGovernance
Tiered pipeline · new market opportunitiesRanked
Tier 1
Tier 1
Tier 2
Tier 2
Tier 3
Tier 3
Impact for the investment →
A structured roadmap — opportunities withheld

How the Babel AI Hub Centralized Workflows and Governance

Kalei successfully deployed the Babel AI Hub, a single governed entry point for all of Babel's AI-enabled workflows, agents, and knowledge. This centralised hub guarantees team-wide adoption, ensures strict data privacy, and features an ROI dashboard to track precise time savings.

Developing AI tools is only half the battle; providing a centralized place for employee adoption is the other half. Today, all of Babel's AI-enabled workflows and agents live behind a single governed entry point: the Babel AI hub powered by Kalei.

This shared repository makes AI accessible to everyone at Babel, empowering the entire agency. Furthermore, Kalei integrated an adoption and ROI analytics dashboard that tracks Babel's usage, time saved, and overall return against initial baselines.

BabelAI HubIllustrative viewPowered byKALEI
HomeStrategyAgents LibraryWorkflowsAI ToolsOpportunitiesKnowledge BaseGetting StartedGovernanceAdoption & ROI
Babel × Kalei
The Babel AI Hub
A single governed entry point: workflows, agents, opportunities, knowledge and governance — easily accessible to everyone.
Workflows
Agents
Knowledge
Governance
Usage
vs. baseline
Time saved
vs. baseline
Return
vs. baseline
The Babel AI hub powered by Kalei · usage, time saved and return tracked against initial baselines · figures withheld
“What's been the biggest benefit so far is the shared context and knowledge that's been brought together by the sharpest minds in our agency. Kalei brings the method and the structure, we bring the craft. Everyone starts at the same high level and has access to this thinking from the Babel AI hub. All while protecting privacy and human judgement with governance and accountability built in by design.”
Ash Lockett · Babel Senior Director, AI Lead

Kept alive

AI is not one and done. It needs to be continuously validated, enhanced and governed. Working with Babel, Kalei provides an ongoing support and continues to optimise workflows and agents as it is used by the team in real life. New workflows and agents are being proposed as the team continues to explore and understand the opportunities of AI-enabled work.

This has only been possible by unlocking a shared standard and understanding from the wider team. Where once verified best practice used to live with one or two people, the team's growing confidence from daily use is feeding a live roadmap.

BabelAI HubIllustrative viewKALEI
Live roadmapFed by the team
Proposed by the team
Proposed by the team
In validation
Optimising
New agent
Validated in real useOngoing
Shared standardWhole team
Continuously validated, enhanced and governed — detail withheld
“Whereas we used to have one or two people feeding into what we could do with AI, we now have the whole team engaged and proposing new opportunities for Babel. AI is still really in its infancy. You learn by doing, experimenting and learning from real world feedback. Our AI roadmap is now defined by those doing the work, carefully curated to deliver maximum impact and unlock human creativity.”
Ash Lockett · Babel Senior Director, AI Lead

What ROI and Efficiency Gains Did Babel Achieve with Kalei's AI Solutions?

Kalei's AI implementation drastically accelerated Babel's proposal generation, reducing the time required to assemble a 34-page draft proposal from two weeks down to a single day. This intelligence automation increased Babel's new business win rate and enhanced existing client campaigns.

Of all the Kalei deployments, Babel's proposal process provides the clearest signal of ROI. By connecting Babel's intelligence process and powering it through AI, Kalei helped Babel uncover sharper insights for B2B PR campaigns.

Metric processBefore Kalei integrationAfter Kalei integration
Proposal generation timeOne to two weeksOne to two days
Time intensive taskDesktop research, competitive analysis and intelligence gatheringHuman creativity and strategic thinking
AI usageFragmented, siloed, individual workflowsCentralised via the Babel AI hub

The AI workflow elevated Babel's overall capability to match the tacit skill of Babel's most experienced strategists. Because Babel's team spends less time on background analysis, it allows the team to flourish in what they are good at - bringing creativity and bravery to B2B PR and marketing campaigns. With more time to fine-tune the ideas and campaigns, Babel has seen an increase in new business wins.

“New business in the agency world is a time intensive process for revenue that's not yet guaranteed. It used to take days if not weeks to conduct all the desktop research and analysis to create the intelligence to come up with a brave creative proposal. Now, the intelligence is gathered in minutes, meaning we can spend more time developing the ideas, winning more work as a result.”
Ash Lockett · Babel Senior Director, AI Lead

But it's not just new clients that are experiencing the benefits. Current clients are also reaping the rewards of Babel's new AI-enabled intelligence process.

“Because we built all the workflows using modular agents, we've discovered different use-cases we weren't necessarily expecting. Within our proposal workflow, our intelligence tools have been able to deliver excellent analysis into our clients' industries and their competitive landscape. This has enabled us to fine tune our PR and marketing roadmaps to deliver bigger impact and better results.”
Babel team

How Kalei continues to support Babel's AI evolution

Babel began as a hands-on engagement and is becoming Kalei's standing capability, one that stays current as the technology changes, rather than being re-solved every time the models move. Workflows and agents continue to be tested and optimised, while their AI roadmap is constantly assessed and verified.

That is the arc Kalei builds toward with every client: from mapping and installing the workflows to keeping them current.

How Kalei Can Close the AI Gap For Your Firm

If your people are already ahead on AI but the structure around them hasn't caught up, and value has not been seen at a company level, and the work is too craft-led to hand to a generic tool, that is exactly the gap Kalei can close.

Talk to us
Order from complexity.