Olly Headey

There was so much more to say about last week's funding announcement than we could fit into a blog, to really express our vision at Tessl to deliver an AI Native Development platform - so I recorded a special podcast episode about it, with Ben Galbraith and Simon Maple. We had a great discussion around: - What is AI Native Development and how does it redefine the traditional software development workflows? - What distinguishes Tessl's approach to AI Native Development from existing AI-powered tools? - Insights into a spec-driven, pluggable, and open ecosystem model - What role do LLMs play in filling gaps in specifications, and how does Tessl address challenges like validation and randomness in AI-generated code? - How does Tessl's platform approach spec generation, decomposition, and system validation? - What are the goals and timelines for Tessl's product release?  👉 Full Episode: https://tessl.co/4fvfXJ4  🍎 Apple Podcasts: https://tessl.co/4fYeuem 🎙 Spotify: https://tessl.co/48XGDjk #AINative #Tessl #SpecCentric #Funding #AINativeDev

121

8 Comments

if you want to learn about how we built the company and what makes it unique, https://lnkd.in/eQS2VRKf is a really really well written article about us by Bill Kerr - better than what I have managed to put together so far!

34

7 Comments

I believe in paying for software. I use lots of it to build & support my habit tracker app “Did You Do it Today?“, and pay for a fair chunk of it. I’m happy to do so, though, as the stuff I pay for really helps me, and it feels good to support the folks & companies building & maintaining it. Software I pay for: • WordPress → Domain name & web hosting, blogging • Bubble → Full stack, no-code development platform (this software is incredible & is the only reason a non-coder like me can build an app like DYDit) • Stripe → Payments processing • Sinch ClickSend → Daily Check-in emailer • SendGrid → Daily Check-in SMS sender • Titan → Customer support emailer • Google Workspace → Spreadsheets, forms, & other emailing • SnagIt → Screen captures & visual designs Software that’s free: • Trello → Idea & task management • Notejoy → Writing • Figma → Visual designs • Microsoft Copilot → AI image generation • Meta AI → AI image generation • Bubble third-party plugins → Visual elements (eg. charts) • LinkedIn → Marketing • Meta Facebook → Marketing • Instagram → Marketing • Twitter → Marketing • Threads → Marketing Holy shit, that’s a lotta software! We use way more in our daily lives than you’d think, and it’s especially amazing how much you use when you’re trying to build a digital business. I’m pretty content with my paid vs. free balance, especially since I could be, in some cases, using a hacked-together free solution for something I pay for. Also, with the way AI is advancing, I’ll be likely be using & paying for even more in the near future. So, like I said, it feels good to pay for software that really helps you. How about feeling good with me by building your Habit over @ https://lnkd.in/e_8nwVwX #dydit #dyditlife

11

2 Comments

Tiers is here! When Ben Vallack and I launched SetSeed as a subscription model we had this clever idea of easy licensing. One site license gave you everything. Wahoo! Yeah, except it was actually a really dumb idea. 🤦 Great for small sites. Terrible when we realised our platform was powering big sites, even government department ones overseas. So… in the spirit of life being one big R&D journey, we’ve rejigged and automated our limit system. Websites can now be deployed based on intended use. Scale up with ease too. Value isn’t being the cheapest in town; it’s about being the smartest.💡 #FoundersJourney #StartUp #ScaleUp #PluginFree

9

3 Comments

Market update; June 2024.   I’m still receiving DMs from developers either out of work or struggling to find something new.   Here’s a bit of a random dump from me, based on chats I’ve had with clients that might help. Definitely not all doom & gloom but I'm being honest also; - Applying to recession proof businesses; If you’re looking for a role, think about industries like health funds/insurance/utilities/superannuation etc; strong industries needed in our society that always have cash (usually). Whereas SAAS/product/Fintech seem to be trying to stay lean.. - CTOs; are telling me they have projects they need to kick off (platform re-architectures etc) but their CEOs aren’t giving them the budgets they need. It’s a case of doing their best with the team they have. This is another reason were seeing less roles advertised. - Doing more with less; There’s more pressure on senior & staff engineers to be able to demonstrate they have a solid understanding of tech but can also deliver at speed, meaning they have to focus on quality AND shipping, which in theory improves profit. Also with less hiring on-shore due to budget constraints, this means a leaner team, meaning existing teams have to deliver. - Junior engineers; To land roles, juniors need to show proactiveness and the ability to get things done. This would position them closer to the expectations of a mid which is what hiring managers want right now (theyre just not telling you). - Consultancies; A few businesses have reduced their local teams, to move to an outsourced model as they minimise perm headcount. This means these consulting businesses are busy and are hiring. So your Accentures/TCSs of the world could be good places to apply if you're looking? - New fin year; is approaching meaning new budgets & usually more roles. However this doesn’t mean from July 1 you’ll see a huge influx of roles posted. One CTO I chatted to has been given the same capex budget as he was for last year (not great). - Fully remote roles; becoming rarer. Most businesses require 2 days a week in office (some, not all). If remote is your jam, try to find work in startups/small businesses wheres theres flex (that said, I do know some corporates contradicting what they’re saying; letting some of their high performing contractors work interstate). - Ignore job boards; Roles are getting filled quickly due to so many candidates on the market & not even making it to the internet. Just because you don’t see a live role in a companies careers page, doesn’t mean they’re not hiring. If it’s a solid business, send them your details. - “I’m applying for roles and not hearing anything” and the likelihood is you never will unless selected for interview. Some companies will be getting 1000 applications a week for 1 role. They cant reply to everyone. Don’t take it personally. - The green banner debacle; it doesn’t make you look desperate. Stop worrying about it 😊 Hit me up if you need to chat; [email protected]

139

33 Comments

Im always interested to find how things are built and really enjoyed reading the case study on Craft Cloud being built on Laravel and deployed to AWS: https://lnkd.in/eAnv5MW8 Hosting millions to scale is an interesting challenge every time but always learn something new. At Framework we have a lot of experience in deploying systems to PHP on Lambda AWS, so if you have any queries or wish to discuss then always happy to discuss. #craftcms #cloud #serverless #php #laravel #aws

11

Am I weird for being excited about seeing this sort of cache hit ratio on the Design Week website? It's the best we've achieved on any site - in part because we give ourselves the budget to work on what we feel is really important on a website. One of those things being lightning fast performance. People don't like waiting to read stuff! The content on most pages is delivered to the browser within around 50ms or less and pages will often be fully rendered (minus images if you have a very large screen) within 200ms. The images part is another interesting one. Rather than serving a few pre-sized thumbnails, because image quality is super important to designers, we serve up images rendered from very large sources at pixel perfect resolutions to our readers. This does result in a little slowness with images on 4k monitors, say, but for the vast majority of users results in optimised bandwidth and really clear presentation. That's the nerdy stuff over. It's not very opinionated though, so I'll not be getting much attention!

16

5 Comments

Throughout the years I have talked to Italian startups (many are impressive) and occasionally have been surprised by companies solely focused on the Italian market, which will seem limiting for a typical US VC investor or entrepreneur, mostly (I speculate) because of the goals of their strategic investors. In the, US the domestic market is large enough to support a single market strategy, and still most companies focus on a global footprint, with rare exceptions (e.g. Zillow only expanded to Canada; OfferUp is US only). In Israel, where the domestic market is tiny, all startups (especially consumer internet startups, e.g. the many free-to-play game companies) are default-global. If you focus on a single market, it'd better be big, because as copycats emerge across the world, catching up to a global footprint may no longer be an option.

9

2 Comments

📣 SENIOR ENGINEER / HEAD OF TECH 🤙 We're looking for a topnotch #seniordeveloper who's ready to grow into a #headoftech to join our #startup. Please share within your networks and have a think. I'm super keen to get a recommendation rather than use recruiters or post on job boards. Must be UK based. JD below. Big thanks y'all! We're looking for a smart, kind, determined and ambitious #seniorengineer to join our smart, kind, determined and ambitous team. If you dream of working at a startup, having equity so you can reap the rewards of your hard work, read on. If you're as obsessed with data as you are customer experience, then get in touch. And if you're passionate about the future of AI, then make sure you get this job! Resale fashion is growing 5x faster than new fashion (+22% yoy). Yet it still takes 3x longer to shop, and over 60% of people don't both due to effort and risk involved. The world doesn't need another PLACE to shop secondhand, but it does need another WAY to shop secondhand. Circle of Style takes the slog out of secondhand, by matching people with product using 1P customer & inventory data, AI/ML. ROLE A #fullstack engineer, who loves data but also passionate about creating incredible customer experiences. Able to set strategy with driving the tech vision for the business, and also get hands dirty everyday. Share options available. Opportunity to work at a venture that: Already has thousands of users and generating income. Innovate UK grant funded (and going for more funding) Part of the Google High Potential Startup Program & Accelerator Walpole Brand of Tomorrow Ahead in terms of AI/ML adoption working with thought-leaders in the field, doing some really exciting work (if you want to learn a lot you will!). World-class award-winning team, as well as advisors including the ex-CTO of IBM Watson. Leadership skills: - naturally commercial and strategic, obsessed with data, and able to prioritise. - good communication skills are key, as you need to communicate with non-tech partners, investors, and a cross-functional team. Information sharing and clear communication is vital. - a pragmatist who knows when to leverage existing software and when to build our own. - obsessed with creating great customer experiences as a B2B2C business. - ambition to do famous work and build a profitable, valuable business, and exit successfully. - ability to work independently and remotely. - ability to manage a team. Experience & Qualifications: - ecommerce or marketplace platforms is helpful, as is knowledge of apparel. - full-stack, Javascript, React, Next.js, AWS and GCP (BigQuery, Looker). - experience and/or strong interest in AI and ML. - passionate about working with data, as we use proprietary data in everything we do - recommendations / matching engine experience a plus. - Bachelor's degree in Computer Science and/or Maths a bonus.

25

4 Comments

🎉 Exciting News! ThirdRetail has secured SEIS advance assurance within 24 hours of self-application! 🚀 🔍 What is SEIS? The Seed Enterprise Investment Scheme (SEIS) is a UK government initiative designed to encourage investment in early-stage startups. It offers tax reliefs to investors, making it easier for young companies like ours to raise capital. 👋 Who are we? ThirdRetail is revolutionizing offline distribution for D2C brands! 🏪 We're the Airbnb of retail shelves, making it affordable and accessible for direct-to-consumer brands to scale their physical presence. Our AI-powered platform removes cross-channel blindspots, providing seamless integration between online and offline retail. 💡 Our mission: - Empower D2C brands to expand offline - Optimize retail space utilization - Bridge the gap between digital and physical retail 🤝 Calling SEIS Funds! Are you interested in being part of the third channel of retail sales? We're currently raising our pre-seed round and would love to chat with SEIS funds looking for exciting investment opportunities. Let's connect and explore how we can shape the future of retail together! 📈 Anil Junagal #RetailTech #SEIS #SEISFUNDS #Startup #D2C #Retail #Marketplace #B2B #SaaS

42

17 Comments

I have some downtime so in between mountain biking, drinking lots of tea and Christmas shopping, I'm having a look at React Native to see if it's as good (or bad) as other multiple platform tool chains I have used. Wish me luck. Any tips ? So far quite impressed with QR code generated in terminal <geek>

4

2 Comments

Every tech founder want cash (nothing new here) 😂 Cash comes in many ways: - Revenues - the best form - Equity finance - Debt - Grants I (https://deallite.uk) sit on investor data with relationships built over a long time. Whilst I (Deal Lite) share this data through monthly subscriptions, I avoid investor introductions at all cost due to a number of reasons. I'm currently evaluating a light weight approach of investor introductions for Techcelerate members. Deal Lite also has the ability to include live fundraisers. Just wondering whether there is sufficient demand for us to create a product related service. It could look like below (not finalised - very early research): Product: - Listing on https://deallite.uk (company needs to keep us up to date so the data remain live) - Pitch deck sent to us for a quick review and inclusion with above. Service: For personal introductions, you would need to follow additional process: - Join Techcelerate as a Member. - Build credibility and trust within our community - Sign an agreement on light weight fundraising - Make introductions Of course, I would chat with Stuart Scott-Goldstone to understand legal ramifications. Whatever decided, it has to be a very light weight product related solution to ensure a decent quality of service. No decision has been taken. Just researching at present. I bet David John William Bailey might have some views on this. #fundraising #investment #DealLite

8

Bit more on what tipped the scales to bootstrapping vs VC for us. A big factor was financial pressures—or the lack of them! 1. Fewer financial costs from infrastructure etc. Back in the day tech businesses were enormously expensive to start—and run—so you couldn't do it on your own. e.g. You probably needed pricey hardware, Oracle databases, etc. That's not the case anymore. Especially when lots of tools offer startup versions or open source. The guts of the Flagsmith platform were essentially free and we scaled to billions of monthly requests on our own budget. That would have cost a fortune 20 years ago. 2. We built out of an agency When we started Flagsmith we were running an agency, so we could just build organically in line with open source growth. Without the financial pressures to build quickly, we didn’t need capital. When we grew we started weighing up VC money. There are a lot of obvious perks to the VC track. The cash injection can give you a massive competitive advantage—and fast. It's tempting when you're in a pool with goliaths. But for us, the obligations just weren't worth it. (e.g. Meetings, externally set targets, reporting, timelines. Erk) Flagsmith's a sustainable business now so fortunately the financial pressures still aren't there. We keep sidestepping the VC path and cash injection in favour of growing efficiently and sustainably... even if it's slower.

45

2 Comments

🌟 This Week's Diary of a CTO! 🌟   Interestingly, since changing my role title to CTO, I've been invited to more events, roundtables, and debating sessions, further feeding my imposter syndrome. Do I really have something useful to say in a room full of more technical people who have worked their way up to CTO? (Hopefully, only time will tell!).   I recently joined a gathering of CTOs and CEOs that looked a bit like Oakland - small and mid-sized consultancies serving the enterprise market. The hot topic? AI....obvs! Initially dubbed the shiny solution in search of a problem, it seems the hype has simmered down, and things are starting to settle, with a market analyst in the room observing that attention is moving away from tactical proof points to more considered strategic adoption of AI, echoing Gartner's recent sentiments.   But here's where I differ. As a CTO, I feel obligated to steer our clients toward understanding AI's true potential and targeting the right challenges. Sure, AI sparks excitement. Can it alter human history? Perhaps. Will it deliver substantial ROI? The verdict's still out. But here's the crux: regardless of the tech, the key lies in pinpointing the right problem.   I believe the goldmine lies within the vast amounts of unstructured data yet to be explored and exploited. Tools like LLMs can enable businesses to elicit insights better. And the growing AI Agent frameworks are a great next step in this space (more on that in future posts!). But this still needs to be fed back into a process to drive a different outcome and ultimately result in a benefit. And you need to know that the process (patient) can adopt the AI (implant). This is the hard bit!    While news of slowing interest rates is a relief, business leaders' pressure to justify IT spending has never been more intense. The stakes are high—miscalculating AI deployment could prove astronomically costly.   So, the next time you're hit with "Why aren't we using AI?" or "AI can handle that, right?" think about: 1 - Do I understand the problem we think it can solve, and is there a benefit if we resolve it? 2 - Can the process we'd implement survive the adoption/inclusion of AI?  3 - If we solve the problem and quantify the benefit, can we communicate that to leaders in a way they'll understand (this is an entirely different post altogether!) It all boils down to clarity—a notion I recently discussed with Greg Freeman in our latest video. Stay tuned for more insights! 😉 #CTO #AI #TechLeadership #DataInsights #EverythingData

32

3 Comments

Pleo I've been using and recommending this firm to my customers since they first arrived in the UK, but no longer. Not only did they take funds from my account without informing me (as a partner, you get a free account but they cancel your "benefits" after 12 months unless you keep referring new customers), I asked to withdraw the rest of my funds over 3 weeks ago, and they still haven't credited them to my bank, yet they are unavailable in my Pleo wallet. They told me standard was 5-10 working days, but can't tell me where my money is. I have been emailing them, and they just give me vanilla responses about it being with "the support team". No person to contact in the support team, no timelines on when I will get a response, no responses unless I chase them (often more than once). Of course, there's no telephone line to call them on, which seems pretty standard these days - not acceptable, just standard. I thought I would look at the FCA to see how to start a complaint, and this came up - rather worrying. Anyone else had problems with them? Think I will be looking for a new solution for my customers now. More upheaval for them, and for us, but I don't like my customers to be exposed to issues like this.

35

34 Comments

Had a realisation about Product People and why we are good at both taking and mitigating risks – we don't have a specialism, so we're great at saying "So What"? This allows product people to be experts at "comparing apples and pears", even when there is incomplete data (which is always the case). This sounds obvious (and maybe it already was to others, apologies if so), but it only just dawned on me. Some examples: 1) A peer in payments has come to you, as they are really worried that PSD2 isn't turned on and they think this is a big risk as it's legally mandated – they want you to make it a top priority to fix it. For them it's top priority as they are accountable for payments, but as a product person who is agnostic, you ask "how much is the fine and how many companies have been fined so far"? They are flabbergasted that you aren't taking it more seriously, but they don't realise you are taking it as seriously as every other risk, but you just aren't as close to it as they are. 2) Technology want to borrow engineers from all the product teams to upgrade the ecommerce platform the company uses, as it's out of service soon and is therefore a "burning platform". For them it's a top priority, as they accountable for service reliability, but as a product person who is agnostic, you ask "what happens if we don't do anything"? The answer is "nothing", but over time the risk of something going wrong and being easily fixable increases. 3) The Brand team want to refresh the look of a product that has performed well at MVP stage, to make it more in line with other parts of the portfolio, but this would slow down the build and scale of the product. For them it's a top priority, as they accountable for the companies brand salience, but as a product person who is agnostic, you ask "what happens if we ship the MVP branding and do you have any data to suggest the new brand would improve our metrics"? And I could go on, it's the same with UX design, engineering, marketing, sales and everything else – all have needs, but none are created more important by default. All will be greeted by product people with a polite "So what", as they are balancing the needs of everyone (including the customer).

17

#Xero new pricing https://lnkd.in/em-XNCxS They have found a way to charge more by keeping the subscriptions more or less the same but adding per item charges. I've never even considered leaving Xero before since I started with them first in 2008/09 (can't quite remember). But I don't like a gun held at my head. What are other options in the market for micro businesses? I don't like the idea of banking and accountancy software being from the same company so won't go with Tide and similar. My very first company (2004 to 2006) used #Sage50 (or whatever it was called) I think. #accountancysoftware

1

A QA engineer walks into a bar. Orders a beer. Orders 0 beers. Orders 99999999999 beers. Orders a lizard. Orders -1 beers. Orders a ueicbksjdhd. First real customer walks in and asks where the bathroom is. The bar bursts into flames, killing everyone. A QA engineer walks into a bar, WITH a team of QA AI agents. Asks AI agents to test the app while drinking an ice-cold Modelo Especial. Trained AI agents order X factor of Y elements in Z universes, in parallel using customer behavior data to identify common patterns in the test coverage. The QA engineer keeps drinking and curates the strategy. Real customer walks in, and shares a Modelo with the QA engineer. Lessons: 1. AI is here to help QAs, not just replace them 2. Modelo Especial es muy deliciosa y refrescante

32

1 Comment