Bikal Tech

The full agenda has been released for tomorrow's Digital Infrastructure North 2026 (DIN26), sponsored by Village Heatshare and Lenovo with support from Autotrader. Last chance to register to attend: buff.ly/H4dJPFP

@Lenovo @DialogXR #DigitalTransformation

It's finally here! Today marks the first Digital Infrastructure North 2026 #DIN26! We are proud to be supporting a new event for the North 🇬🇧 bringing together those passionate about the infrastructure and connectivity that serves the public sector. Convened by Manchester Digital, Cooperative Network Infrastructure, Greater Manchester Combined Authority and LINX, today’s event brings together leaders senior leaders from government, infrastructure operators, technology companies and research organisations to explore opportunities and shape the future of digital infrastructure in the north. Our very own LINX CEO Jennifer Holmes, EMBA kicked things of by sharing an insight into why the north matters for a resilient internet in the UK! #LINXManchester #PeeringandMore #Interconnection #NetworkControl #Data #KeepTrafficLocal #NetworkSecurity #Peering #LINX #IXP

UK founder starter pack (save this): → incorporate via Companies House: £100, 30mins → bank: Mercury (US)/Starling (UK) → legals: SeedLegals (~50% new UK startups) → EMI BEFORE first hire → HMRC advance assurance pre raising → cap table tools from day 1 a weekend, <£1k

@number10gov being run like @SpursOfficial.

@eliegsaliba1 @b1rdmania It looks like an AI wrapper, for small claims court applications in the UK.

@b1rdmania @BIKAL_TECH It would hit regulatory hell as long as the output of the system is fuzzy. LLMs are great at determining user intent which is great when you have a specific outcome in mind. Garfield.ai does a great job of providing the outcome (retrieving unpaid invoices).

Legal AI is where fintech was in 2015. Everyone's selling to banks. Nobody's built the Monzo yet. Two days at a legal AI conference in Lisbon. $50M raises for contract review CRM tools Anthropic ships in a product update. 18-month sales B2B cycles into law firms where tech-averse partners make the final call. The gap is internal at law firms (180 day product cycles and sign off from 60yr+ partners 🥵) or 5.5m UK SMEs who can't afford a solicitor for most stuff and get nothing. Gonna ship some open source legal infrastructure this week. Early days. Wltm London solicitors bored in their current role, not tech-phobic and open to jamming on some ideas.

@b1rdmania @eliegsaliba1 It is for small claims, these are usually self service in the UK. It looks good but it is automating a simple process, and looks like a openai or gemini plugin.

@eliegsaliba1 @BIKAL_TECH Garfield law is a nice model. CTO / tech co and then solicitor leading the law firm and fully SRA regulated.

Introducing the newest member to the team, @DialogXR XR Ltd. An enterprise grade #sovereign #AgenticAIwill be assisting our team in operating the business operations. This is completely on-premise through the deployment of a @Lenovo #SR630 V4. instagram.com/reel/DX8HctDB2…

@Jameswise @thetimes true say.

The UK is already seeing the benefits of being an AI hub. But we’re not doing a good enough job of explaining the opportunity ahead. I wrote for @thetimes today on being an ambitious nation again - and the need to fight the AI doomers. We had similar opportunities in the internet era, but failed to seize on them. We can’t miss again.

@manishamishra24 At which minute does he mention the 5,000 #GPU infrastructure that they spent $500m on to run this? Please let me know.

Jane Street hired a junior researcher on a compensation package reportedly ranging from $220K to $600K after he built systems that used AI to process and analyze massive-scale datasets. Not millions. Trillions of data points. In a 1-hour session, he walks through the exact setup behind it. How he uses machines to scan impossible amounts of information. How AI helps surface patterns humans would never catch manually. And why the people winning right now are the ones learning to work with data at insane scale. Most people use AI for tweets and homework. Meanwhile others are using it to build research systems that would’ve required entire teams a few years ago. Skip Netflix tonight. Watch this instead and study the workflow.

@AthenAlgo @manishamishra24 And 5000 GPUs. No one is mentioning that...

@manishamishra24 This is exactly the shift nobody is talking about. It's not about knowing more than everyone else anymore. It's about building systems that can process what humans physically can't. Trillions of data points. One researcher. A few AI pipelines. That's the new leverage. 🔥

@coreyganim Is automation of the same old workflows is an optimal use of #AI? This will be disrupted by actual AI that will be finding exponential value in the business.

A guy offered me $1,000 to follow him around for a day and tell him where to implement AI in his business. I didn't take the deal (because this is not scalable at all) but it did make one thing clear: business owners are desperate for AI guidance. What they don't need is a developer or an "AI automation" agency. They need someone who can simply tell them where AI fits into their existing workflows. I turned this into a product called the AI Assessment. I sell them for $999. Here's the exact system you can copy: THE OFFER $999 for a custom AI report. 3-7 specific tool recommendations tailored to their workflows. Guarantee: 5 hours returned per week minimum, or 100% refund. Sounds aggressive but it's really not. Average ROI I've delivered is 6+ hours/week saved for ~$40/month in tools. STEP 1: THE 45-MIN INTERVIEW Google Meet call recorded with Fathom. Questions I ask: -Walk me through a typical day -What tasks do you hate but can't delegate? -Where are you the bottleneck? -What do you wish you could automate but haven't figured out? Goal: surface their pain in their own words. Not what they think they should automate, but what actually eats their time. STEP 2: FEED THE TRANSCRIPT TO CLAUDE I've built multiple Skills that chain together for this step but if you're just starting out a simple prompt will still work: "Analyze this transcript of a business owner describing their workflows. Identify the top 5-7 areas where off-the-shelf AI tools could save them significant time. For each, recommend specific tools and estimate weekly time savings." Claude catches patterns you'd miss. Connects pain points across the whole conversation. I review, add my own picks, cut anything that feels like a stretch. STEP 3: BUILD THE REPORT IN GAMMA Structure: -Executive summary (their top 3 opportunities) -Impact/Effort matrix (focus on quick wins) -3-7 tool recommendations with implementation details -4-day quick wins plan -What comes after quick wins (preview of upsells) -Financial impact -Next steps Looks like a $5K consulting deliverable but only takes me 5 minutes from template. STEP 4: THE 30-MIN REVIEW CALL This is where trust gets built and upsells happen. Screen-share and walk them through each recommendation. Gauge their appetite for help. About 40-60% want help implementing. That's where the real money lives. THE UPSELL LADDER Process Optimization: $3-5,000 No-code automation (Zapier/Make): $1-3,000 Custom GPTs: $3,000+ Custom AI Skills: $3-5,000 Full agent implementation: $5-10,000 + $1K/mo retainer Real example: business brokerage owner gets 400 emails every time he lists a new business. Every email asks the same 4 or 5 questions. For $3,000 I built him a custom GPT trained on each listing's marketing package that he can send to potential buyers. Eliminated 95% of his email workload per listing. WHY THIS WORKS Business owners don't need another AI course. They don't need to learn prompt engineering. They don't need a developer. They need someone to look at their specific situation and tell them exactly what to do. What we're selling here is clarity. HOW TO START 1) Know the tools. Spend a month using Claude, ChatGPT, Fathom, Gamma, Zapier. 2) Do 2-3 free assessments to build your process and get testimonials. 3) Start charging. $499. Then $749. Then $999. Raise every 3-4 clients until people say no. 4) Build your upsell menu (where the real money is made) Go out and make it happen.

@0xMovez Ya, and the access to the 5000 GPUs they use to run this.....?

Jane Street AI Engineer revealed how they trained their own LLM for trading to make $22.5B/year 16 minutes. free. straight from tier-1 quants. bookmark & watch - this is the most honest "AI inside a hedge fund" talk ever published. forget the "AI trading bot" YouTube grifters. This is the real inside view: data, training, evals, integration. then start building your own bot using post below.

@thejayden They have 5000 x #GPUs brav. That is probably more important than the quants.

And that’s why quants have salaries starting at $500k+ Here’s one of the engineers explaining how they use LLMs to generate $30B+ every year.

@pvergadia but in the cloud... no #sovereignty

Palantir's end-to-end Agentic AI Architecture!

@51bodila Great, thanks. Will you also send me the credit for the 5,000 GPUs that they have access to as well please?

Jane Street Head of Technology showed the code that generates $13B profit - using it, you can build your own hedge fund (ofc with ai) you’ll watch 50-min of "vibe coding" from a person who has worked for 24 years at a tier-1 fund bookmark - it will change the way you see trading systems

Dialog XR is the antithesis of public #AI tools. instagram.com/p/DX4M0tsmtGW/…

@mogulinfluence A simplistic take on where #AI tech can actually go. Chasing the $500 tax return filing service problem statement is an under utilisation of the technology. There are larger and more significant services that have greater value than software services that AI can tackle. @DialogXR

Read this if you want to know what YC and Sequoia are actually betting on in AI space right now:

@51bodila The video won't explain how New York-headquartered trading firm Jane Street has amassed a massive computing fleet of over 5,000 powerful #GPUs for high-frequency trading and algorithmic execution. You can build your own with $500m in compute.

Jane Street AI Engineer explained how the hedge fund uses LLMs to make $20.5B/year - they trained their own model that generates profit for them 16-min and you'll learn how tier-1 funds use AI in their calculations Bookmark & watch - no matter what, this is the best AI + Hedge Fund lecture out there. Then start build your own hedge fund or bot on @Polymarket