One more part of the BPM Skills series is here!

I am very happy to share with you inspiring answers from two more experts. Below you can learn more about the role of process automation and customer experience in modern BPM.

As always, you can either read everything or use the navigation below. Enjoy!
Boris Krumrey
Steve Towers

Boris Krumrey

Boris Krumrey, the Global VP Automation Innovations at UiPath, is responsible for driving the UiPath Automation Innovation agenda to transform organisations with Agentic AI for customers and partners, showing the art of the possible with AI. Boris invented and runs the UiPath Innovation Labs, which he describes as the “Agentic Automation Kitchen” to inspire businesses exploring new customer and work experiences. In his initial role as Chief Robotics Officer at UiPath, he led the product roadmap and the integration design for RPA and AI technologies.

WWW: LI profile

1) How do AI and other trends impact the way organizations manage and run their processes?

AI is pushing organizations from “process flows” toward end-to-end work systems that combine deterministic orchestration + adaptive agentic work.

• Deterministic orchestration remains essential for enterprise scale: clear execution semantics, auditability, predictable outcomes, and governance.
• AI agents expand what can be automated beyond structured tasks into interpretation, drafting, summarization, classification, exception handling, and guidance under ambiguity. But more importantly depending on the selected model and context grounding capability AI agents expand to reasoning, planning, analysis and making decisions.
• Multi-agent systems increasingly break linear process/case thinking: an orchestrating agent coordinates work while specialist agents handle different stages of a case and can reorder, revisit, or escalate steps based on context—more like humans managing real-world work.
• This increases the need for an orchestration layer (e.g., UiPath Maestro) to keep adaptive behavior inside guardrails and measurable outcomes, rather than letting execution become freeform and hard to govern.
• A parallel trend is “prompt-only workflows” (natural language specs in markdown, etc.). This can be useful for prototyping but often struggles in enterprises due to governance, auditability, scalability, and LLM cost/latency.

2) What are the skills, techniques, behaviors, and attitudes that can help BPM practitioners create value for their organizations in 2026?

Core BPM skills stay relevant—but the role expands. In 2026, BPM value comes from combining process discipline with agentic capability and operational reliability.

Skills & techniques

• Agentic design fundamentals: defining agent goals and constraints, choosing tools, designing grounding/context (RAG) plus memory, guardrails rules, setting confidence thresholds and escalation rules, and evaluating outputs.
• Orchestration-first thinking: designing for exception paths, retries, compensations, human-in-the-loop approvals, and evidence/auditability.
• Multi-agent and adaptive case patterns: understanding how orchestrating agents coordinate specialized agents across case stages.
• Case management / CMMN-style thinking: modeling non-linear, event-driven, situational work that doesn’t fit a strict flow.
• BPMN as execution backbone: BPMN remains important in enterprise automation; it provides a deterministic, inspectable model.

Behaviors & attitudes

• Prototype quickly, operationalize deliberately: experiment fast but insist on governance, observability, and measurement before scaling.
• Outcome orientation: prioritize measurable business impact over perfect modeling artifacts.
• Governance-by-design mindset: treat safety, compliance, and accountability as design inputs, not after-the-fact additions.

Important forward-looking point: to model agentic work better, BPMN likely needs to evolve (or be extended) with an “Agentic Task” concept that visualizes what’s inside the agent step: allowed tools, grounding sources plus memory, guardrail rules and internal decisioning/escalation gates.

3) What are the best resources to learn those skills?

UiPath resources (hands-on, enterprise-relevant)

• UiPath Academy – Agentic Automation Associate Training
  https://academy.uipath.com/learning-plans/agentic-automation-developer-associate-training
• UiPath Academy – Build your first agent with UiPath Studio Web
  https://academy.uipath.com/courses/build-your-first-agent-with-uipath-studio-web
• UiPath Academy – The UiPath Maestro collection (BPMN + orchestration)
  https://academy.uipath.com/learning-plans/the-uipath-maestro-collection
• UiPath Community Edition (to experiment)
  https://docs.uipath.com/overview/other/latest/overview/product-download

Broader BPM resources

Well, this community would know best

4) Which skills are no longer relevant or not practically applicable yet (hype)?

A few things to be cautious about:

• “Prompt-only BPM” / fully freeform agent execution as the default: great for
  experimentation, often not enterprise-ready for governance, auditability, predictability,
  and cost/latency at scale.
• Fully autonomous agents for core regulated processes: still more aspiration than
  default; hybrid patterns (deterministic backbone + bounded agentic tasks) will
  dominate.
• DMN as the universal answer for decisioning: DMN won’t disappear, but in many
  cases decisioning becomes hybrid—rules where precision matters, agents/AI where
  judgment under ambiguity is needed. So DMN may be used more selectively or in
  combination.
• Notation purity without operationalization: focusing on diagram perfection while
  ignoring execution reliability, exceptions, evidence, and measurement is increasingly
  unhelpful.

Jump to the top

Steve Towers

Steve is passionate about helping people and businesses transform to better ways, with happier lives. Whether that is individuals, teams or companies I apply proven and tested ways from the very best individual and corporate achievers to help you codify your own success, happiness and future.

Named one of the 30 most influential Global Customer Experience Experts in 2022. An experienced business transformation leader with over 40 years of success in driving and achieving organizational goals in both the private and public sectors in a variety of key ‘C’ leadership and top-level consulting positions. Recognized across industries including Business Process Management, Enterprise Architecture, Customer Experience and Lean Six Sigma

WWW: Website
WWW: LI profile

Which BPM skills will be hot in 2026? (Outside-In is the divider)

You’ve heard this story.

Office full of clever people.
Screen flipped to show process map.
Someone declares. “We just need to automate it.”
And then customer reality walks through the side door.
“I had to chase you”
“I had to repeat all of that information”
“How could this take three weeks?”

That’s when BPM chooses between staying in the weeds, or becoming a competitive performance discipline.

Because in 2026, hot BPM skills aren’t about fancy new diagramming techniques.

They’re about designing operations that actually deliver customer outcomes, at speed and scale, with control.
That’s why Outside-In is never been more relevant.
And why I keep coming back to the CEMMethod.
Not as another buzzword.
As the practical foundation that keeps improvement efforts from devolving into internal theatre…

2026: BPM matures (at last)

Automation is now the easiest thing. AI assistants abound. Low-code tools are ubiquitous.
That’s great…if your focus is replacing humans.
But most businesses aren’t just trying to automate away people.
They’re trying to create customer experiences that make buyers happy, drive loyalty, and encourage upsell.
As work gets automated from top to bottom, those who rely on manual effort will get crushed.
They’ll either automate like crazy…
Or they’ll fail to change.
Here’s the new game:
1) Outside-In process architecture (don’t redesign processes, design experiences)
2) Process knowledge (use data to understand current-state performance)
3) Simplify aggresively (design experiences with humans, bots, and AI working together)
…and more!

…but let’s reality check. In theory theory works but in practice theory fails.

I’ll walk you through three examples that bring new game skills to life.
…and one punchline on why BPM + Outside-In = everything coming together.
Ready?

1) Outside-In process architecture (experiences first, processes second)

Questions you ask as an Inside-Out designer:

• “How do we run this department faster?”
• “How do we serve this product more efficiently?”
• “How do we automate this task?”

Questions you ask as an Outside-In designer:

• “What experience are we trying to create for the customer, start-to-finish?”
• “What is the Successful Customer Outcome?”

Example: insurance claims

The old way

One process force fits the customer experience.
Treatment goes from triage, to assessment, to repair, to payment.
The customer calls for an update. Rinse and repeat.
No single group owns the full experience.
Everybody focuses on their internal process.
The customer gets lost in the shuffle.

Solution: Outside-In design.

You might have multiple underlying processes internally.
But you design around one unifying claim experience.
And the internal workflows adapt to support it.
Compare that to the “customer experience” outlined above.
Faster? Sure.
Empathetic? Not exactly.
But here’s the thing.

Outside-In processes are not “softer”.

They are actually stronger.

Why? Control.

When you’re solving for the entire experience instead of your tiny process slice, you build complete ownership at the top.

• Hiccups get the attention they deserve.
• Customers aren’t handed off willy nilly.
• This is a desired outcome at every interaction.

To put people first, you need supple experiences.
That means exception handling built into the fabric of how work gets done.

2) Process knowledge (stop guessing, start measuring whats-going-on-actually)

By 2026, workshops aren’t enough.
Sure, you can hold a meeting and decide what you “think” is happening.
But the best BPM teams use data from the actual execution environment to show what is really happening.

• Delays.
• Exceptions.
• Rework loops.

Example: bank account onboarding

It feels like the credit check is the slow step.
Until you use process information to show where the pain is actually happening.
Customers are mailing in their documents…over and over.
They can’t read your instructions.
So they call, get clarification, and start all over again.
So you fix up the document request.
Simplify language, reduce ambiguity, clarify steps.
Then automate the steps that can’t be messed with.
Outside-In the secret sauce:
Elapsed and cycle time matters.
…but so does customer effort and experience predictability.
If you don’t know what your customers see, you don’t know the whole story.
So measure the process from their POV too.
Connect it back to your customer account health scores so you can act.

3) Orchestration (putting people, automation, and AI together without destroying reliability)

Holy automation, Batman.
2026 will not be framed as a battle of “BPM vs AI”.
Oh no.
BPM will become the control layer that enables AI to be safe, measurable, and scalable.

Example: contact centre
AI can write first draft replies.
Summarize customer calls for agents.
Rock your world.

But what happens if it starts autonomously triggering claims?
Escalations?
Shipping orders?
Suddenly your AI has multiplied your downstream errors by a factor of…
MACHINE SPEED.
When you’re designing processes that work WITH AI (and not merely for IT), reliability is king.
That means defining the process states.
Making decisions explicit.
Guiding AI to only do the predictable “messy middle”.
And letting people handle the high-risk exceptions.
Scales like nothing else.

4) Decision intelligence (making “why” auditable again)

As soon as AI enters the conversation…
CEOs and boards want to know one thing:
“Explain that decision to me.”
Get used to it.
That means decision modelling is about to become a red-hot skill.
Not fuzzy concepts of “policy”.
Clear definition of decision logic that can be tested, managed, improved.

Example: public sector service eligibility

Someone either gets approved or declined.
But when you’re dealing with people’s livelihoods, you need to be able to tell THEM why.
AI can help you sort documents.
Flag potential risk points.
Augment processing power.
…but the actual decision needs to be clear, logical, and above reproach.
Outside-In the secret sauce:
People chase you when they don’t understand the what and the why.
Stop failure demand from happening in the first place.

5) Process observability (management by fact, at scale)

Observability will shift from being a “nice to have” reporting feature…
To a foundational skillset.
Built into process design from day zero.
Not fancy dashboards hacked together after your app is finished.
Signals and metrics you can use to see process health.

• Detect drift.
• Catch failure modes.
• Reduce rework.

Example: retail returns
“You improved handling time by 30%!”
…but customers are MORE irritated than ever.
Why? You only measured how fast your employees work.
“We’re expediting returns but no one is getting resolved.”
“They keep transferring me to another department.”
Don’t let this be you.
Measure what customers see.
Connect back to your service experience metrics.
That’s CEMMethod in action:
closing the loop between customer experience and process improvement.

6) Experience-led standardisation (only standardise where it helps)

Experience-led standardisation is the holy grail.
Standardise where consistency adds value.
Protect flexibility where the customer experience matters most.

Example: healthcare patient discharge
Yes, some things need to happen every-single-time.
…but not everything.
So you create a standard way of working that provides consistency.
But you bake in intentional flexibility for your agents to handle exceptions.
Throw away the pillowcases.
Corridor CPR doesn’t scale.
…but built-in flexibility does.

7) Change activation (because your workers gotta use it)

You know what doesn’t exist?
Processes that no one follows.
Upload your fancy new process to the SharePoint portal.
Or…
Guide your workers at the moments of truth,

• right there in front of the customer,
• on the device they’re already using.

I’ll let you guess which drives actual change.
– – –

Conclusion

When’s the last time you saw BPM done right?
Operations execute flawlessly.
…and customers complain about nightmarish experiences.
Don’t choose BPM as a SILOED activity.
Choose BPM + Outside-In thinking.
Connect your operational efforts to the front-line experience.
…and own the entire experience from end to end.
Like we just talked about.

Simple 10-skill checklist for 2026 BPM:

1. Outside-In experience-to-process design
2. Process mining & process intelligence
3. Process observability (signals that drive behaviour, not vanity metrics)
4. Orchestration design across people, bots, and AI
5. Decision modelling & decision governance
6. Exception management & “casey” thinking
7. AI for BPM (because it helps…some. And sometimes it hurts.)
8. Experience-led standardisation (when to standardize, when to flex)
9. Change activation & adoption strategy
10. Value storytelling (connecting process efforts to customer outcomes + the triple crown)

My new book “Everyone Loves Great CX – Your Customer Experience Playbook” is now available. We review the themes here and turn them into an Actionables.

Learn more here: https://bpgroup.org/everyoneloves

Jump to the top