James Futhey

Now that we’ve delivered custom workforce management solutions for a few contractors in California, I’ve had a chance to really see the folks in the field and the management issues that trace back to the days when laptops were a luxury thing. (I’m talking about small to mid-sized organizations here.) It turned out to be quite a conservative vertical, with many still relying on outdated stuff. Challenges Faced by Field Service Businesses: A survey by Verizon Connect highlights the most challenging issues: - Customer satisfaction (56%) - Profit margins (48%) - Talent acquisition and retention (47%) Other challenges include: Technology Alignment: Many companies use multiple outdated technologies that don’t integrate whatsoever. Operational Cost Increase: Dispatches generate revenue but also generate crazy overhead. Talent Pool: Poor technology frustrates field staff, leading to higher turnover. When it comes to field service management, some industries are particularly notorious for poor management. The top five industries that struggle the most with FSM are: Telecommunications: Companies in this sector often face challenges with scheduling, dispatching, and maintaining the vast network infrastructure, leading to delays and inefficiencies​ Utilities: The management of services like electricity, water, and gas supply is complex, often plagued by outdated systems and equipment, leading to frequent service disruptions and customer dissatisfaction​ Healthcare: Field service in healthcare involves the maintenance and repair of critical medical equipment. Issues here can severely impact patient care and safety due to delays in service and equipment downtime​ Construction: Managing field services in construction is challenging due to the dynamic and fragmented nature of projects. Poor coordination and communication often result in project delays and increased costs​ Manufacturing: This industry deals with maintaining machinery and equipment both in-house and at customer sites. Ineffective FSM can lead to prolonged downtimes and significant productivity losses​ TBC

23

I bet the public officials I hold accountable really flarking hate me. Especially when I tell them I’m working with AI, which I recognize still makes mistakes regularly, & then give them a flurry of codes, which I do need to verify, that show they do need to do what they said they’d do when they got the job. When you priss off officials with information we haven’t had easy access to until now, you’re doin’ it right. https://lnkd.in/g3PPzv9Y If you aren’t using language models regularly yet, get on that! We’re at a point like when television, the home computer, & the internet were introduced. It’s only going to get more ingrained in our lives every day that passes. It’s not always right but the batting average is between 750 & 950 which is a huge lead when starting projects. & just having the ability to ask in plain language for information that has been a huge chore in the past is worth it. Think about skills that will become required & basic in the not too distant future. Now if you’ll excuse me, I’m off to verify this information, along with a lot of other stuff I’ve dealt with today.

1

Starting to see some sensible reflections on Gen AI at last. The key word to remember here folks is 'Rework'. Rework is like not putting the right torque into the bolts of your engine. You drive for a while and everything seems fine.. Then there's a loud crunching noise! You have to pull apart the motor, fix up the damage from the lost bolt and start again. Checking everything is done right this time. That halt happens at a time when you really didn't want it. Like when a customer checks the references and notices that the GPT made one up (something a human might be fired for).. Forcing you to check everything. Re-do everything. Meanwhile the technology companies are saying "Don't worry about hallucinations! We'll fix that with a gazzillion GPU's.. " Team, you don't have to worry about hallucinations. There are only hallucinations. We imprint intelligence when a snappy combination of words comes through. It's not knowledge. Own your own biases and use GPT's as a super search engine. "Give me ten ways to.." "Describe the nuances of this system behaviour.." "Tidy up the grammar in this paragraph.." "Start me off with a list of ideas to..." Places where we start to enter dangerous territory.. "Summarise the meeting discussion.." <in a meeting that you didn't attend> "What were the key points of <literature>.." "Which is the best candidate for <insert job>.." "Imagine you're a really competent director, with deep ESG and compliance experience, now assess this document.. " < I mean what the hell?? > Gen Ai can be useful, but remember it doesn't 'know' anything. It's super handy when it comes to thinking about what we could do.. But AI can't compete with humans around deciding what we should do. AI will never be responsible and doesn't care if it's right or wrong. It will cheerily write up anything it thinks could work. It has infinite patience and no ego to bruise when it gets things wrong..

5

2 Comments

Our team demo'd something on Friday that blew my socks off. Codename: Project Raptor. (Because if you can't have awesome code names for software projects, what are you even doing???) So...what is it? About a year ago I put together this slide to help illustrate what the future of strategic finance might look like. The idea was pretty simple: Today, even with a great toolkit, humans have to do most of the work of *thinking* up the scenarios to run, then using their tools to run those scenarios, then comparing to outputs to find the winners. Tomorrow, you'll wake up to an email from a finance agent of sorts who has been busy all night running scenarios on your behalf with the latest data and cherry-picking the best ones for you to consider putting into action. Project Raptor makes this vision *very* real. In the same way you might be using ChatGPT to write a great draft of your blog post, you'll be using Project Raptor to build a great draft of your next Annual Budget, along with all the sensitivities and cases you might want to explore. It's going to be freakin' awesome. Can't wait to share more soon...DM me or drop a comment if you're interested in getting involved as a beta tester :) HUGE shoutout to Derek Y. the entire Drivepoint engineering team for blowing me and the rest of the team away.

28

✨ Understanding AI for Business Decisions ✨ Are you based in the Columbia River Gorge? 🏔 I'm looking at offering an in-person 2 hour course in Hood River, OR on how AI works and how you can use that knowledge to make educated business decisions. 👩‍💼 Not only will you get this fundamental knowledge, you'll also be in the room with an expert who can answer your questions in real time. 🧠 Often, I'm brought in as a consultant and one of the first questions I get is "What's the difference between AI and ML?" 🤖 Whether you are just starting out thinking about leveraging AI or you have a team experimenting, it benefits you to know how the fundamental technologies work. With the right knowledge, you'll have all the power you need to know if an idea is worth pursuing, if you need to bring in an expert, or if you should go a different direction. 👍 A little about me: 🤖 8 years in the AI/ML industry with companies like Intel, SightLine Applications, Voxel51, and more ✏ 10+ years teaching with organizations like Portland State University and more 📜 Holds a masters degree in computer science specializing in artificial intelligence and machine learning 👩‍💻 Co-Founder and CTO of new AI startup Aequa Interested? 🤔 Fill out the form below! If there's enough interest, courses will be scheduled and you'll be the first to know how to sign up! https://lnkd.in/giByF3NQ For more information, check out www.leilahk.com/courses 👀 #courses #inperson #interested #ai #ml #business #whatisai #whatisml #hoodriver #columbiarivergorge #tech #womeninstem #womeninsteam #kanedaconsulting

11

3 Comments

To those concerned about SB1047, especially those like Arvind Narayanan, below, who don't think it's possible to have “reasonable assurance” of safety, I think it's critical to understand regulatory law and how legislative language is interpreted, and what a term means legally, instead of trying to interpret the phrase yourself. In this case, it's not unclear what "reasonable assurance" means - it has a clear legal meaning related to how audits in other domains are conducted. That is, it means following best-practices and standards for having third-parties do an audit, with sufficient access to conclude that their assessment is not misleading. In AI, practices have been demonstrated for red-teaming and safety audits, including by firms doing internal safety testing, and external audit by groups like METR, as well as standards for safety testing which firms are committing to voluntarily, and practices that are being requested by the US and UK AISIs for audits. What the phrase means is doing those types of activities, not meeting some dreamed-up impossible standard! (This is especially clear given the text as amended on May 16th clarifying that "'Reasonable assurance' does not mean full certainty or practical certainty." But this was unnecessary, since, again, the term is already defined legally in audit and other places to means exactly that!) Moving forward, these classes of methods are being codified by National Institute of Standards and Technology (NIST), the UK AI Safety Institute, and the ISO - International Organization for Standardization's working groups. Once recognized standards exist, these will be legally sufficient for determining whether an audit complies with SB1047's "reasonable assurance" standard - but I agree that until then, there's going to be some degree of uncertainty about what to include in an audit to check safety, and in practice, lawyers will probably suggest that they need to be inclusive of various best practices to be sure they fulfill the requirement.

7

Owners: Are you afraid of hiring the wrong person? ☠️ Use contracting as a way of vetting candidates, when possible. Years ago, we hired an engineer who seemed great, until he started working with us. 👿 Then, toxic behavior began to emerge, like: • Being unwilling to compromise • Talking down to teammates • Ignoring feedback We had no choice but to let him go and restart our search, costing us time, money, and morale. 😓 Missing red flags in the candidate process taught us to, “try before you buy.” 🚩 Taking the time to make sure both sides are a good fit is critical to building a strong team. Besides, it is easier to wrap up a contract than to lay someone off. #atlanta #entrepreneurship #business

5

2 Comments

Owners: Are you afraid of hiring the wrong person? ☠️ Use contracting as a way of vetting candidates, when possible. Years ago, we hired an engineer who seemed great, until he started working with us. 👿 Then, toxic behavior began to emerge, like: • Being unwilling to compromise • Talking down to teammates • Ignoring feedback We had no choice but to let him go and restart our search, costing us time, money, and morale. 😓 Missing red flags in the candidate process taught us to, “try before you buy.” 🚩 Taking the time to make sure both sides are a good fit is critical to building a strong team. Besides, it is easier to wrap up a contract than to lay someone off. #atlanta #entrepreneurship #business

3

No silver bullets... - great stories about how [insert company] does product, often fail to communicate the deep context that helped make those things work - overarching principles that exist in companies deemed high performing, are by definition context-free (but boy do the principles make you nod your head). They *also* lack actionable details. - equifinality: there are often multiple ways to reach the same end-point - what works, stops working - what didn't work, might start working - experience and skills help, but all experience is contextually situated All to say that at the end of the day you need to be a lifelong learner, curious, willing to challenge your assumptions, be willing to try things, be willing to discard them, and find other lifelong learners who are willing to take that journey with you

192

13 Comments

Let’s face it, we’ve all seen those basic chat widgets on websites that pop up and say, “Leave your contact info, and we’ll get back to you.” The user fills out a form, hits submit, and then waits... and waits some more. Or maybe you don't see one at all. In today’s fast-paced world, that just doesn’t cut it anymore. Customers want answers now, not later. And if your response is delayed, there's a good chance they've already moved on to your competitor. A basic chat widget is like a glorified contact form—it collects data but doesn’t actually engage with the visitor. Maybe it will auto populate some FAQs from your website. The difference with the AI-powered assistant we offer at Grow Local is night and day. Instead of just capturing contact info and promising to follow up, our AI assistant engages instantly with leads, answers their questions, qualifies them in real time, and even books appointments directly into your calendar—all while keeping the conversation going. Our AI assistant doesn’t just wait for someone to reach back out; it actively nurtures and interacts with potential leads, offering personalized responses and guiding them through the process with intelligent, human-like interactions. If you're still relying on a basic chat widget that only collects info and says, “We’ll get back to you,” you're missing out on real-time engagement with your potential customers. With Grow Local’s AI assistants, your business is always available—responding instantly, qualifying leads, and even scheduling appointments for you while you sleep. These assistants can come in many forms, including voice agents that people can interact with on your website. Texting agents that respond immediately to a customer that fills out a form, or even one that answers the phone for you. *Most likely, your web developer does not know how to build this for you, but we can provide everything they need to add it to your website. In the competitive world of lead generation, an AI-powered assistant does more than just collect leads—it converts them. Don’t let outdated tools hold your business back. Interested in upgrading your lead generation from passive to proactive? Let’s chat about how an AI assistant can help you capture, engage, and convert leads more effectively. Book your FREE demo here: https://lnkd.in/gfFM5nDg #AIinBusiness #LeadGeneration #BusinessAutomation #Chatbots #AILeadConversion

1

Is your development team slowed down by the number of items they currently have in progress? It's hard to identify what all is in progress when the work is invisible. That's why I highly recommend finding a visualization for your team's WIP. I've worked with many organizations to map out their workflows and WIP, and jaws drop every time when the visualizations reveal how much work is currently in progress. And that's the thing about WIP—the more work items you have in progress, the slower each individual item becomes. That's why visualizing it can be a powerful part of speeding up your team. Listen to this full episode at the link in the comments, or find Agile Bites on your favorite podcast platform. #agile #agility #agilesoftwaredevelopment #techleadership #WIP

22

5 Comments

Use Signal. Encrypt your private communications. If you're using AI, be sure to turn off 2-way sync or memory to protect yourself from data scraping. “Signal either works for everyone or it works for no one. Every military in the world uses Signal, every politician I’m aware of uses Signal. Every CEO I know uses Signal because anyone who has anything truly confidential to communicate recognises that storing that on a Meta database or in the clear on some Google server is not good practice.” https://lnkd.in/eKfFfyzk

5

No action made with the intention to help others is too small to make a difference. I've been thinking about this a lot lately and reflecting on how much potential power I've wasted in the past by not taking any action because I felt any action I could take wouldn't make a big enough impact to matter. Every action you take no matter how small, that challenges a status quo you don't like, is a form of activism. If you don't like what you see in the tech industry today, don't wait for someone else to change it. Use your strengths to be the antidote, one tiny action at a time.

5

Upwork cut their workforce by 21% and saved $60M yearly with automation every company can cut FTEs by 10% or more the problem with automation? it’s not the tech—it’s the know-how how do you use AI to eliminate bottlenecks? here’s what I would advise Hayden Brown at Upwork: ➡️ start with job postings: which roles are they hiring for? what tasks are repetitive? these are your bottlenecks. btw, just spotted “reporting obligations” in their latest posting—a highly automatable task ➡️ prepare your team: transparency is essential. everyone needs to understand the shift, to focus on creativity over routine ➡️ audit processes: observe what employees do, pinpoint what’s essential but automatable, and automate it this is within reach for any business start by checking job postings then audit automatable tasks we’ve created an AI to find every automatable task in your organization DM me to learn more ;)

5

A week ago, a bipartisan bill was introduced called “Enhancing National Frameworks for Overseas Critical Exports Act” or ENFORCE Act (gotta love these acronyms). The bill amends the Export Control Reform Act of 2018 (ECRA) and will give the Commerce Department’s Bureau of Industry and Security (BIS) the authority to control the export of AI and other national security-related emerging technology. Given the ever-escalating AI race with China, the US export restrictions on AI chips, and in light of the recently announced deal between Microsoft and G42, a UAE-based AI firm, the US government does not want to bleed AI technology and would bar its citizens from developing AI systems if national security is at stake (that's precisely what the bill does). Per Reuters, the Microsoft-G42 deal is "an agreement between the two companies that requires each to give security assurances to their respective home governments, but there is no direct agreement between the U.S. and UAE governing the transfer of sensitive technologies." Interestingly, there is not much information on the deal as far as the components of AI technology to be exported. If we are talking about model weights the ENFORCE Act will come in handy.

15

1 Comment

“Blend and Send” is quickly replacing “Post and Pray” The new trend is applying for 1000 jobs while you sleep. Cool. I call this Stuff the Q! Sites like LazyApply and AiApply have customized the point of entry so you now hit a button and it synthesizes: your stuff with the job stuff and blends and sends the stuff into the stuff place at each company. The modern ATS is not built to do what comes next. All of a sudden all the stuff looks just like your stuff - like a perfect match. Hmmmm. First comes stuff. The comes paralysis. Then comes meltdown. It seems we “may have” lost the plot. What do you think?

27

17 Comments