Latest / Chief in Tech Deep Dive Shortcast / Chief in Tech S1E40 - Sabrina Rethakin : Unlocking Tech-Driven Efficiency – A Blueprint for Future Success

Transcript

Chief in Tech S1E40 - Sabrina Rethakin : Unlocking Tech-Driven Efficiency – A Blueprint for Future Success · Chief in Tech Deep Dive Shortcast

• 0:00
  Right now, as you listen to this, there is a
• 0:02
  very high probability that your digital workspace
• 0:05
  is, well, an absolute disaster. Yeah, that is
• 0:08
  almost a guarantee at this point for most people.
• 0:10
  Right. I mean, we are looking at a Harvard Business
• 0:12
  Review study today, and it highlights a statistic
• 0:15
  that is genuinely hard to wrap your head around.
• 0:17
  It really is. It says the average knowledge worker
• 0:20
  switches between 1 ,200 different applications.
• 0:23
  And that's not total apps over a career. That
• 0:25
  is 1 ,200 context switches between apps every
• 0:29
  single day. Which is just a staggering level
• 0:31
  of cognitive friction. I mean, think about it.
• 0:34
  Every time you toggle from, say, a project management
• 0:37
  dashboard over to an email thread. Or like dropping
• 0:40
  into a messaging channel. Exactly. Or checking
• 0:42
  a spreadsheet. Every single time you do that,
• 0:45
  you incur a cognitive penalty. Your brain literally
• 0:48
  has to recalibrate, and that friction just slowly
• 0:51
  bleeds your momentum dry over the course of the
• 0:53
  day. And usually, you know, when organizations
• 0:56
  notice this loss of momentum, the default reaction
• 0:59
  is to treat the workforce like a massive roaring
• 1:02
  engine. Right, just push harder. Exactly. When
• 1:05
  output drops, management just forces the engine
• 1:08
  to run faster. They push the pedal to the floor,
• 1:11
  they demand more hours, shovel more coal into
• 1:14
  the furnace, and basically just hope the whole
• 1:16
  system doesn't rattle itself to pieces. Which
• 1:18
  is a completely exhausting way to work. It really
• 1:20
  is. And if you are listening to this, you probably
• 1:23
  know the exact feeling of being trapped on that
• 1:25
  modern hamster wheel. And, you know, the brute
• 1:28
  force approach to productivity, it works right
• 1:30
  up until the point the engine actually explodes.
• 1:34
  a significant portion of the modern workforce,
• 1:36
  we are hovering very, very close to that explosion.
• 1:39
  Which is terrifying. It is. Actually, as we start,
• 1:42
  you'll notice the backdrop behind me has shifted
• 1:44
  to a cityscape constructed entirely of architectural
• 1:47
  blueprints. Oh, yeah, I see that. That visual
• 1:50
  is very intentional because the new paradigm
• 1:53
  we are exploring today treats workplace efficiency
• 1:56
  not as a metric of speed, but as a structural
• 1:59
  design challenge. Which perfectly brings us to
• 2:01
  the mission of today's Deep Dive. We are looking
• 2:04
  to rescue you from that exhausting cycle by analyzing
• 2:07
  some incredible insights from Sabrina Ritaken.
• 2:11
  She's a tech -driven solutions leader over at
• 2:13
  Walmart. She has some brilliant perspectives
• 2:15
  on this. She really does. Through her recent
• 2:17
  presentation and a related article on sustainable
• 2:19
  efficiency, her central argument is basically
• 2:22
  that we need to stop pushing the pedal harder.
• 2:25
  Okay, let's unpack this. What if true efficiency
• 2:29
  is about stepping back, turning off the machine,
• 2:32
  and entirely redesigning the engine from the
• 2:34
  ground up? So you don't actually have to push
• 2:36
  that pedal so hard in the first place? Exactly.
• 2:38
  That is the fundamental shift Ruth Hackett advocates
• 2:40
  for. Modern efficiency has evolved far beyond
• 2:43
  just operational spreadsheets and time tracking.
• 2:46
  It is a sustainable, highly tech -driven, and
• 2:49
  intensely human -centered design challenge. But
• 2:52
  to build that sustainable engine, organizations
• 2:54
  first have to confront a massive blind spot,
• 2:57
  which is the over -tooling trap. Oh, the overtooling
• 3:00
  trap. The illusion that just buying a new piece
• 3:03
  of software is going to magically fix deep operational
• 3:07
  problems? Yes. The tech solves everything, Meff.
• 3:10
  It feels like having a kitchen stuffed with 50
• 3:13
  highly specialized, incredibly expensive gadgets.
• 3:17
  You know, you've got the avocado slicer, the
• 3:18
  automated egg peeler. The Bluetooth -enabled
• 3:20
  meat thermometer. Yes. But because you have accumulated
• 3:24
  so many disjointed gadgets, you have absolutely
• 3:27
  zero counter space left to actually just chop
• 3:30
  an onion. Aren't these tools supposedly making
• 3:33
  us faster? You would think so, but they just
• 3:36
  end up stealing our space. And what's fascinating
• 3:38
  here is how the source material quantifies that
• 3:40
  exact trap. Oh, right, the poll she did. Exactly.
• 3:44
  During her live presentation, Rietackin ran a
• 3:46
  poll asking the audience to identify their biggest
• 3:48
  barriers to efficiency. And the top answers were
• 3:51
  incredible. I bet they didn't say we need more
• 3:54
  software. Not at all. The audience did not point
• 3:56
  to a lack of automation or a shortage of cutting
• 3:58
  edge tools. The number one barrier was confusing,
• 4:02
  outdated processes. And immediately following
• 4:04
  that was a total lack of clarity and alignment
• 4:07
  across teams. So the problem isn't a lack of
• 4:10
  shiny gadgets in the kitchen. The problem is
• 4:12
  nobody knows what recipe we are cooking. Right.
• 4:15
  And the instructions we do have are from a decade
• 4:17
  ago. Yeah. So if you take a terrible, complicated
• 4:19
  process and buy a brand new enterprise -grade
• 4:23
  software platform to run it, you haven't actually
• 4:26
  solved anything. You have literally just digitized
• 4:29
  your dysfunction. That is exactly it. You are
• 4:31
  executing the wrong tasks just a fraction of
• 4:34
  a second faster. You're just scaling inefficiency.
• 4:36
  Precisely. So the immediate question becomes,
• 4:39
  well, how do organizations actually start to
• 4:41
  untangle this mess? Since throwing more apps
• 4:44
  at it just makes that 1200 -switch nightmare
• 4:46
  even worse. Exactly. The sources outline a clear
• 4:49
  framework for this, starting with automating
• 4:51
  the repeatable. This requires genuinely auditing
• 4:55
  your workflows to identify the mundane, purely
• 4:58
  repetitive tasks. Like basic data entry. Data
• 5:01
  entry, basic quality assurance checks, things
• 5:04
  like that, and then handing those specific isolated
• 5:06
  functions over to technology. Right. And to build
• 5:10
  on that, the framework suggests moving beyond
• 5:12
  just those isolated tasks and looking at the
• 5:15
  broader picture, which leads to the second strategy.
• 5:18
  orchestrating workflows. Which is usually where
• 5:20
  things break down. Oh, completely. Because orchestrating
• 5:23
  workflows means getting different software platforms
• 5:25
  to actually share data. But legacy systems are
• 5:28
  just notorious for being stubborn. They really
• 5:31
  are. If orchestrating workflows is like building
• 5:34
  bridges between isolated islers of data, how
• 5:37
  do we stop legacy systems from burning those
• 5:39
  bridges down? I mean, if our modern apps speak
• 5:43
  a contemporary language, our legacy databases
• 5:45
  speak an ancient dialect. And they simply cannot
• 5:48
  communicate. So we end up with humans manually
• 5:51
  copying and pasting data between the two. Acting
• 5:53
  as human bridges. Exactly. And the human bridge
• 5:56
  is the most expensive and error -prone bridge
• 5:58
  you can possibly build. Absolutely. So to solve
• 6:00
  that language barrier, the sources emphasize
• 6:03
  adopting an API -first mindset. For those less
• 6:06
  familiar, an API functions like a universal real
• 6:09
  -time translator between different software applications.
• 6:12
  Right. So instead of building isolated silos,
• 6:14
  you demand interoperability from day one. Yes.
• 6:18
  You ensure that your inventory system can seamlessly
• 6:21
  hand off data to your compliance system without
• 6:25
  a human ever having to click export. And you
• 6:28
  combine that with the third strategy, which is
• 6:30
  prioritizing data. Establishing a robust data
• 6:33
  strategy with clear ownership is vital. You simply
• 6:36
  cannot optimize a workflow if you cannot measure
• 6:39
  its outputs. That's a fundamental rule. You can't
• 6:41
  optimize what you can't measure. Right. And when
• 6:43
  organizations actually integrate these three
• 6:46
  strategies, automation, orchestration, and data
• 6:48
  prioritization, the results are profound. They
• 6:51
  really are. The Walmart compliance domain example
• 6:53
  in our sources is just the perfect illustration
• 6:56
  of this. Oh, the compliance workflow is a stellar
• 6:58
  case study. They had this legacy process that
• 7:01
  required 200 different touch points. Wait, 200
• 7:03
  touch points for one process? 200. Think about
• 7:06
  the friction of that from the worker's perspective.
• 7:09
  A human being had to navigate 200 individual
• 7:12
  interactions. That's insane. Opening emails,
• 7:15
  downloading PDFs, cross -referencing spreadsheets
• 7:18
  against this massive compliance matrix, seeking
• 7:21
  approvals all just to get one single task done.
• 7:25
  it took weeks to complete. It sounds agonizing.
• 7:28
  I mean, that is the literal definition of the
• 7:30
  hamster wheel we were talking about. Completely.
• 7:32
  But rather than just buying a new compliant software
• 7:34
  package off the shelf and hoping it fixed things,
• 7:37
  they deeply analyzed the actual bottleneck. Because
• 7:40
  the friction was in the gaps between the tools,
• 7:42
  not the tools themselves. Exactly. They realized
• 7:44
  the friction was in the manual cross -referencing
• 7:47
  and data extraction. So they built a targeted
• 7:50
  machine learning model specifically designed
• 7:52
  to parse those unstructured documents. So it
• 7:55
  automatically extracts the relevant data. Yes.
• 7:57
  And it validates it against the compliance rules
• 7:59
  in the background. The mechanism of that ML model
• 8:02
  is what drives the efficiency here. Wow. Because
• 8:05
  it handles the extraction and validation instantly,
• 8:07
  it only flags the for human review. It shrank
• 8:11
  a 200 -step nightmare down to a simple two -step
• 8:13
  process. Just two steps. Two steps. The system
• 8:17
  processes the data and the human reviews the
• 8:20
  final output. What used to take weeks was reduced
• 8:23
  to two hours. weeks down to two hours. And the
• 8:25
  best part is it didn't require firing anyone
• 8:28
  or forcing people to work through the weekend.
• 8:30
  It was just applying the right technical mechanism
• 8:32
  to this specific bottleneck. Right. And the true
• 8:35
  organizational value here is the compound interest
• 8:38
  of time. Oh, I like that phrasing. By saving
• 8:41
  weeks of labor on that single compliance task,
• 8:44
  those teams didn't just clock out early. They
• 8:46
  reinvested that capture time into analyzing and
• 8:50
  optimizing other areas of the business. That
• 8:52
  is how a sustainable engine begins to actually
• 8:55
  power itself. That makes so much sense. So once
• 8:58
  an organization does this hard work, orchestrating
• 9:00
  their data, eliminating those massive legacy
• 9:03
  nightmares, what does the next frontier of this
• 9:05
  look like? Because the horizon of what is possible
• 9:08
  shifts dramatically. It really does. The sources
• 9:11
  point towards six future trends that are actively
• 9:14
  reshaping how we operate right now. First, we
• 9:16
  are moving away from automating single tasks
• 9:18
  and entering an era of hyper automation. Right,
• 9:21
  hyper automation. Which takes the logic of that
• 9:23
  compliance example and applies it to entire life
• 9:26
  cycles. Instead of just automating, say, an onboarding
• 9:29
  email, you automate the entire end -to -end journey.
• 9:32
  The data gathering, risk analysis, account provisioning.
• 9:35
  All of it connected seamlessly. And driving that
• 9:38
  is the second trend, a fundamental shift in human
• 9:42
  AI collaboration. AI is finally shedding its
• 9:46
  reputation as a mere replacement for human labor.
• 9:49
  It is increasingly positioned as a co -creator.
• 9:51
  So less like a machine taking over the factory
• 9:54
  floor and more like a highly capable colleague
• 9:56
  brainstorming alongside you. Exactly. And because
• 9:59
  these models are becoming so adaptable, we're
• 10:01
  seeing the third trend, which is personalized
• 10:04
  productivity. Tailoring it to the individual.
• 10:07
  Right. The AI isn't just a generic company -wide
• 10:10
  chatbot anymore. It learns your specific workflows,
• 10:12
  your individual use cases, and tailors its assistance
• 10:15
  to how you uniquely manage your time. Which obviously
• 10:18
  accelerates how fast people adopt it. It does.
• 10:20
  But to scale these capabilities across an entire
• 10:22
  enterprise, you rely on the fourth trend, the
• 10:25
  democratization of technology. OK, the no -code
• 10:27
  tools. Yes. Through no code tools and generative
• 10:32
  AI platforms like Claude, everyday users with
• 10:35
  zero traditional programming background can now
• 10:38
  build complex applications and custom workflows
• 10:41
  in a matter of minutes. Okay, hold on. Here's
• 10:43
  where it gets really interesting because I want
• 10:46
  to push back on that concept of democratization.
• 10:48
  Okay, go for it. You just noted that anyone can
• 10:50
  use an AI to code a custom app in two minutes
• 10:53
  without knowing how to program. Doesn't that
• 10:56
  risk creating millions of new tools and plunging
• 10:58
  us right back into that 1 ,200 app nightmare
• 11:00
  we just talked about? That is a very real risk.
• 11:03
  Right. If every department manager is suddenly
• 11:06
  empowered to build their own bespoke software
• 11:08
  solutions, aren't we just flooding the kitchen
• 11:10
  counter with millions of new, highly customized,
• 11:13
  disjointed gadgets? This raises an important
• 11:16
  question, and it actually highlights the precise
• 11:18
  trap the source material warns against. Democratization
• 11:22
  without strict discipline is just accelerated
• 11:24
  chaos. Accelerated chaos, exactly. Giving everyone
• 11:27
  a hammer does not build a house. It usually just
• 11:30
  results in a lot of broken windows. This is why
• 11:32
  the sources stress the vital importance of intentionality.
• 11:36
  You have to use the technology to solve a pre
• 11:38
  -existing critical business problem. Instead
• 11:41
  of working in reverse, what often happens when
• 11:44
  people get these powerful no -code tools is they
• 11:47
  build a shiny, fun, new application just because
• 11:49
  the AI makes it easy. Right, and then they wander
• 11:51
  around the organization looking for a problem
• 11:53
  they can fix. Exactly. Working in reverse is
• 11:56
  the enemy of efficiency. It really is. And to
• 11:59
  prevent that chaos, organizations must enforce
• 12:02
  the fifth trend, responsible AI adoption. This
• 12:06
  means establishing clear guardrails around privacy,
• 12:09
  fairness, and security, ensuring that when citizen
• 12:11
  developers build these tools, they are scaling
• 12:14
  with trust. That makes a lot of sense. And finally,
• 12:16
  the sixth trend, which might be the most important.
• 12:18
  It requires a profound cultural shift. Efficiency
• 12:21
  is no longer just an IT initiative. It requires
• 12:24
  a total transformation in how cross -functional
• 12:27
  teams design, decide, and deliver work collaboratively.
• 12:30
  Which brings us to the most critical failure
• 12:33
  point of this entire ecosystem, the human beings
• 12:36
  on the ground. Always the hardest part. Truly.
• 12:39
  Because you can have the most perfectly governed,
• 12:41
  hyper -automated, beautifully orchestrated architecture
• 12:44
  in the world, but it is entirely useless if the
• 12:48
  workforce refuses to engage with it. The human
• 12:50
  factor is where the vast majority of technological
• 12:53
  transformations just meet their end. Technology
• 12:56
  only provides the capability. human adoption
• 13:00
  provides the actual value. Yeah, and the Walmart
• 13:02
  presentation shares a highly relatable story
• 13:05
  regarding this exact tension. They were rolling
• 13:08
  out a new AI -based inventory management system
• 13:11
  for their store associates. Right. The objective
• 13:13
  was straightforward. Help the associates understand
• 13:16
  inventory levels in real time and eliminate a
• 13:18
  lot of manual checking. But the associates didn't
• 13:20
  understand the underlying mechanics of the technology.
• 13:23
  They didn't know how the AI was arriving at its
• 13:25
  conclusions. Exactly. And consequently, they
• 13:28
  actually pushed back and resisted the rollout.
• 13:31
  Which is an entirely logical reaction. The presentation
• 13:34
  notes a core psychological truth here. People
• 13:37
  simply cannot comply with things they do not
• 13:40
  understand. Yeah. If a black box algorithm suddenly
• 13:42
  dictates your daily tasks and you have no visibility
• 13:45
  into its reasoning, the natural human reaction
• 13:48
  is suspicion and resistance. You feel a loss
• 13:51
  of agency. I mean, if someone walked into my
• 13:53
  kitchen, installed a smart fridge, and told me
• 13:56
  the fridge was now going to dictate exactly how
• 13:58
  and when I chop onions, I wouldn't thank them
• 14:01
  for the efficiency. No, you'd hate it. I would
• 14:03
  unplug the fridge. So how did Walmart actually
• 14:05
  navigate this resistance? Because they didn't
• 14:08
  just mandate compliance from the executive suite.
• 14:10
  No. They pivoted to a strategy of co -creation.
• 14:13
  The leadership team went down to the store floor
• 14:15
  and actually sat with the associates. They gathered
• 14:18
  candid feedback. listened to the specific points
• 14:20
  of friction, and began iterating the application
• 14:23
  alongside the end users. So they ran localized
• 14:26
  test trials. Yes. And they incorporated the feedback
• 14:29
  in real time. They demystified the black box
• 14:31
  by making the associates active participants
• 14:33
  in the design process. They turned the users
• 14:36
  into the architects. Exactly. But to pull off
• 14:39
  that kind of rapid iteration based on user feedback,
• 14:43
  the underlying technology has to be flexible.
• 14:45
  If you are going to co -create, you cannot use
• 14:48
  rigid monolithic software. Right? This is where
• 14:51
  the concept of composable architecture becomes
• 14:54
  essential. Composable architecture is the structural
• 14:57
  foundation for co -creation. Okay, composable
• 15:00
  architecture. Think of traditional monolithic
• 15:03
  software like pouring a concrete mold. Once it
• 15:07
  sets, it is incredibly difficult and expensive
• 15:09
  to change. If an associate has a brilliant idea
• 15:13
  for an improvement, you essentially have to break
• 15:15
  the concrete and pour a new mold entirely. Whereas
• 15:18
  composable architecture is more like building
• 15:20
  with Lego bricks. You have these modular API
• 15:23
  -first components that can be easily swapped
• 15:25
  out, rearranged, or updated. Exactly like Lego
• 15:28
  bricks. So if the human feedback suggests a certain
• 15:30
  workflow isn't intuitive, you just unclip that
• 15:33
  specific Lego brick and snap a new one into place
• 15:35
  without tearing down the entire building. Precisely.
• 15:38
  And that structural flexibility must be paired
• 15:41
  with cultural flexibility, specifically with
• 15:44
  the sources defined as psychological safety.
• 15:47
  If you are asking your workforce to experiment
• 15:49
  with new tools and co -create solutions, you
• 15:52
  have to cultivate an environment where teams
• 15:54
  have the autonomy to test things out. They need
• 15:57
  to understand the boundaries, and crucially,
• 16:00
  they cannot face penalties if an experiment fails.
• 16:03
  Yeah, the concept of failing fast gets thrown
• 16:05
  around a lot in corporate jargon, but genuine
• 16:08
  psychological safety means actually supporting
• 16:11
  the failure. It does. It means not firing someone
• 16:14
  when a new automated process glitches during
• 16:16
  a trial run. It requires establishing clear guardrails
• 16:19
  so the failures are educational, not catastrophic.
• 16:22
  Furthermore, it mandates continuous upskilling.
• 16:25
  We cannot expect a workforce to suddenly collaborate
• 16:27
  with AI without providing the necessary support
• 16:29
  system. We have to train them. Right. Organizations
• 16:32
  must invest heavily in helping their current
• 16:34
  employees acquire these new technical and analytical
• 16:37
  skills while simultaneously rethinking the profiles
• 16:40
  of the talent they hire for the future. So what
• 16:42
  does this all mean for the listener as you navigate
• 16:45
  your own workflows? I think the overarching narrative
• 16:48
  of our sources today proves one undeniable point.
• 16:52
  Technology is a multiplier. It is not a crutch.
• 16:55
  That is a great way to put it. It cannot fix
• 16:58
  a toxic culture, and it cannot magically repair
• 17:01
  a convoluted, outdated process. It simply multiplies
• 17:04
  the reality of whatever you already have. If
• 17:07
  your processes are aligned and clear, AI will
• 17:10
  multiply that efficiency. If your processes are
• 17:13
  chaotic, AI will just help you generate chaos
• 17:15
  at an unprecedented speed. Well said. To synthesize
• 17:19
  the actionable takeaways for untangling that
• 17:21
  chaos for everyone listening, first, start small.
• 17:24
  You do not need a multi -million dollar enterprise
• 17:26
  budget to begin automating wisely. Right. There
• 17:29
  are accessible, free tools available right now
• 17:31
  that can eliminate your most mundane, repetitive
• 17:33
  tasks. Exactly. Second, build for change. Adopt
• 17:38
  that composable modular mindset we talked about.
• 17:40
  Design your systems and your teams so they can
• 17:42
  adapt to future variables rather than locking
• 17:44
  yourself into a rigid concrete mold based on
• 17:47
  today's limited vision. And most importantly,
• 17:49
  empower your people. Make the end users an integral
• 17:53
  part of the design process from day one. If we
• 17:56
  connect this to the bigger picture, the definition
• 17:58
  of efficiency has fundamentally evolved. It is
• 18:01
  no longer a localized exercise in squeezing a
• 18:04
  few more drops of output from an exhausted workforce.
• 18:06
  Not at all. It is a strategic, holistic design
• 18:10
  challenge. Our objective is to engineer systems
• 18:13
  that operate smarter. certainly, but also operate
• 18:17
  more humanly. We are stepping back and redesigning
• 18:20
  the engine so the ride is actually sustainable,
• 18:22
  but As we look toward this incredible hyper -automated
• 18:25
  horizon, there is one provocative thought I really
• 18:28
  want to leave you with. Something to ponder as
• 18:30
  you audit your own 1 ,200 daily app switches.
• 18:33
  OK, let's hear it. If hyper -automation eventually
• 18:36
  handles all the mundane, repetitive tasks, the
• 18:38
  200 touch points, the data entry, the basic quality
• 18:41
  assurance, and AI steps fully into the role of
• 18:44
  our constant high -level co -creator, what happens
• 18:46
  to the entry -level jobs? The foundational tasks
• 18:49
  where knowledge is actually built. Exactly. The
• 18:52
  tasks where junior employees traditionally learned
• 18:56
  the ropes. If the AI executes all the boring
• 18:59
  grunt work that historically taught us how the
• 19:01
  business actually functions on a granular level,
• 19:04
  how will tomorrow's workers build the foundational
• 19:06
  expertise they need to eventually guide and govern
• 19:09
  that AI? That is a fascinating paradox. Right.
• 19:12
  We desperately need to redesign the engine so
• 19:14
  we don't have to push the pedal so hard, but
• 19:16
  we still need to ensure the next generation of
• 19:18
  drivers actually understands how the engine was
• 19:20
  built. It's a lot to think about. Thank you so
• 19:23
  much for joining us on this deep dive. Stay curious,
• 19:25
  take a hard look at your digital workflows, and
• 19:27
  we will catch you next time.