Mohneesh Patel

Autonomous Marketplaces for Commercial Vehicles Ecosystem Photo showing the Vyngo team at a dealer’s site in Navi Mumbai, Maharashtra, India. From left to right, Atul Batra, Aparna Anand, Vikas Agarwal, Ravi Padaki, Adwait Vichare & Sachin Kumar. For the past decade, digital marketplaces have evolved from listing platforms to transaction engines. They helped buyers find sellers, enabled payments, and layered in basic trust. AI is about to push marketplaces into something far more powerful. In India, there are an estimated 8–10 million truck drivers. A significant percentage of them aspire to become owners — even if it is just one truck. But today, the journey from driver to owner is filled with friction: ·        Limited visibility into fair vehicle pricing ·        Uncertainty around vehicle condition ·        Informal negotiations with power asymmetry ·        Financing eligibility revealed late in the process ·        60–70% reliance on NBFCs, often at higher interest rates ·        Fragmented insurance options ·        Paperwork that feels intimidating For a not so educated driver, buying a truck is not a transaction. It is an income engine decision. A wrong move can destabilize everything. Now imagine something different. An intelligent agent that works for the driver 24x7. ·        Understands his earnings history ·        Simulates EMI feasibility based on route-level data ·        Benchmarks fair pricing ·        Negotiates within smart thresholds ·        Compares financing options in real time ·        Optimizes insurance ·        Coordinates documentation He does not just “buy a truck.” He steps into ownership with clarity. That is not convenience. That is economic mobility. Dealers as Regional Liquidity Engines India’s used commercial vehicle market is dominated by hundreds of thousands of regional dealers. They are not just sellers — they are liquidity engines for local economies. Yet most operate with: ·        Fragmented demand signals ·        Manual price discovery ·        Working capital stress ·        Uncertain financing outcomes Inventory often sits idle and capital remains locked. Now is an opportunity to scale capability with autonomous agents. By analysing regional freight patterns, demand signals, transaction benchmarks, and financing conversion data, they enable dynamic pricing and faster inventory rotation. Even a 10–15% reduction in holding time can dramatically improve returns in capital-intensive categories like trucks. Dealers move from reactive selling to AI-amplified portfolio management. Technology does not replace the dealer. It augments him. At Vyngo, we are building towards an AI-native mobility platform designed to evolve into what I term "autonomous marketplace" infrastructure for regional economies. Not merely enabling transactions but enabling ownership at scale. And here’s raising a toast to the audacity of the truck driver becoming a maalik. Atul Batra Vikas Agarwal Aparna Anand Vyngo

81

7 Comments

I've been studying how Garry Tan — President & CEO of Y Combinator — shipped 3 production services and 40+ features in 60 days. Part-time. While running one of the most important institutions in tech. His 2026 run rate is 810x his 2013 pace. Same person. Same brain. Different infrastructure. He open sourced the whole thing. It's called gstack. Here's what it actually is — and why engineering leaders need to pay attention: Most AI coding tools give you a better autocomplete. gstack gives you a virtual engineering team. 23 specialists, all slash commands, all composable: → /office-hours — a YC-style founder interrogation before you write a single line → /plan-ceo-review — challenges your scope like a real CEO would → /plan-eng-review — locks architecture, data flow, edge cases, tests → /review — finds the bugs that pass CI but blow up in production → /qa — opens a real browser, clicks through your app, files bugs, fixes them → /ship — syncs main, runs tests, opens the PR The sprint runs: Think → Plan → Build → Review → Test → Ship → Reflect Every step feeds the next. Nothing falls through the cracks. Here's the compression ratio table that should change how you staff: +-----------------------------+------------+-------------+---------+ | Task | Human Team | With gstack | Savings | +-----------------------------+------------+-------------+---------+ | Boilerplate / scaffolding | 2 days | 15 min | ~100x | | Test writing | 1 day | 15 min | ~50x | | Feature implementation | 1 week | 30 min | ~30x | | Bug fix + regression test | 4 hours | 15 min | ~20x | +-----------------------------+------------+-------------+---------+ What changed? → The tooling finally matches the ambition. → When a single engineer can run 10-15 parallel sprints simultaneously — each with its own planning, implementation, review, and QA — the constraint is no longer headcount. → The constraint is judgment. → That's always been the constraint. We were just paying to paper over it with bodies. I've been thinking about what this means for engineering leaders specifically. Your team's output isn't capped by the number of engineers anymore. It's capped by how good your process is. A 10-person team with gstack or a similar agentic workflow can outship a 50-person team without it. The numbers are that stark. The tool is free. MIT licensed. No premium tier. No waitlist. If you lead an engineering org and you haven't looked at this — you probably should. It could clear the biggest bottleneck in your team's shipping velocity. #SoftwareEngineering #AItools #AI #EngineeringLeadership #Productivity #ClaudeCode #Startups #FutureOfWork #AgenticAI

21

1 Comment

While I’m over here crafting "𝙘𝙖𝙩𝙘𝙝𝙮" LinkedIn hooks and fluffy strategy pieces, the actual geniuses at Pinch AI are busy engineering 𝟵𝟬% 𝗽𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲 𝗴𝗮𝗶𝗻𝘀. They just moved our feature runs from hours to minutes. 🚀🚀 🦀 I’ll stick to the emojis. 🦀 If you want to see what real technical heavy lifting looks like (and why I leave the data architecture to people far smarter than me), check out Mandy's (Mandeep Kaur) breakdown on our switch to DuckDB.

12

When we started ZEPIC, we had one belief ✨ Every business is sitting on more context than it knows what to do with. The problem was never the data. It was bringing it all together in one place 🌐 So we built a platform. One that could connect any system, capture any signal, and give businesses a complete, real-time picture of everything happening across their org ⚡ Then we went further. We built AI infused automation on top of it. So, businesses could connect systems and channels and act on that context automatically. Not just see what was happening. Do something about it. That worked. And it pointed clearly at what needed to exist next. The next layer wasn't another feature. It was agents, purpose-built to close every gap an enterprise has. Not just on top of ZEPIC, but powerful enough to stand alone. Bring any ambition to life. Own outcomes, not just tasks. Introducing Neura 🚀— the first Agent HQ built to close AI gaps in enterprises. Three things we got right: Real-time context at the core. Every Neura agent runs on a living context layer with every customer signal, every touchpoint, always current. Agents that know your business the way your best people do. Agents that work as a team. Build your entire agent organisation inside one HQ. Teams across functions, shared memory, inbuilt orchestration. Agents that don't just execute, they own outcomes together. Built to match your ambition, whatever it is. One agent or an entire workforce. Augmenting your team or fully autonomous. Human in the loop or end-to-end control. Internal-facing, customer-facing, or both. Your infrastructure, your LLM, your rules. Neura meets you exactly where you want to go — and scales with you from there. Neura amplifies ZEPIC. ZEPIC powers Neura. Transform your AI ambitions into a connected Agent Workforce. Take a look → www.neurahq.ai 🎉 Bharathi Kannan Naveen Sreelesh Roheet Divya Srithar Arun Kumar Christopher Nikil Santhosh Aditya🙌 Grateful to the entire team for the amazing effort behind Neura. #EnterpriseAI #AIAgents #Neura ZEPIC

66

4 Comments

Congratulations to the team at Emergent & Mukund Jha on crossing $100M ARR. That milestone is a clear signal that AI-native application building has moved from experimentation to serious economic impact. We are entering a new phase of software development. The first wave was about velocity. Turning ideas into working applications in days instead of months. Reducing friction between intent and execution. The next wave is about governed scale. As Exterview AI - Agentic Talent Intelligence Platform (MFS 26) builds with strict enterprise Azure Complaince based platform aligned with ISO 42001 for AI governance EU AI Act high risk readiness NIST AI RMF and NIST CSF ISO 27001 and SOC 2 Type II Regional privacy laws such as GDPR, DPDP, and CCPA We are seeing firsthand how different enterprise architecture becomes once you operate in regulated environments. Large database schemas. Hundreds of APIs. Bias detection and Responsible AI pipelines. Fine tuned models. Agentic ReAct (Reasoning + Acting) Extensive middleware orchestration. Deep observability. P95 and P99 latency commitments. System reliability, uptime commitments  SLA Contracts Speed still matters. But trust, auditability, and governance become equally critical. One exciting opportunity for platforms like Emergent is enabling startups building enterprise systems to combine rapid application generation with governance ready patterns. Exportable compliance blueprints, infrastructure aware deployment options, and enterprise control hooks could unlock a powerful new segment of founders building regulated, mission critical software. The next decade of AI software will belong to platforms that balance velocity with governed scale. Congratulations again to the Emergent team. Inspiring to watch this category mature in real time.

18

1 Comment

We built an AI system that outperformed Claude. On evaluating creative portfolios, which is what Mela's vetting layer runs on. 48% exact match rate against human evaluation ratings as ground truth. MAE of 0.52 versus Claude's 0.81 on the same portfolios. Every score within one point of the real answer. This matters because evaluating creative work is one of the hardest things to automate. Design and creativity don't have right answers. They have judgment calls. Getting AI to that standard is a different problem than evaluating code or summarizing text. We're rolling out Mela's beta in two weeks. The vetting layer is what the whole trust model sits on, and this is what makes it work. The reason it beats Claude isn't that it's smarter. It does one thing. Built for one task, calibrated against real human judgment, with no generalist noise getting in the way. Last week I wrote that generalist loses to specialist. We just proved it on ourselves. If you're a freelancer, brand, or agency interested in being part of the beta, DM me.

9

In this week of WoodenScale AI's Founder of the week - We are glad to have Nithya Jagadeeswaran, who is building FinPulz. FinPulz is a modern financial platform designed to help everyday investors make smarter, safer investment decisions through curated financial products, structured learning, and trusted banking partnerships. Founders & investors, if this resonates with what you're building or investing in, feel free to connect with Nithya directly and check out FinPulz

15

1 Comment

Nancy Chu shared her learning with us in our last community session and summarized them here in substack article. Key takeaways for me were: - Product Sense is non-negotiable in the age of AI — not just for PMs, but for engineers as well. - Tools like Vibe (Lovable/V0) or Velocity (Claude Code) are meant to demonstrate your product sense and execution ability — not just your ability to build. - As you move beyond L6, you’re expected to consistently demonstrate strong product sense, execution excellence, and leadership across all interview rounds. - Past L6 (especially in AI roles): Interview questions are less abstract (e.g., “design a fridge for the blind”) and more grounded in what you have actually shipped. Even at MAANG-level companies, discussions center more on real-AI impact from your CV and ownership. Pls watch, read or listen- but don't miss. https://lnkd.in/gGVziRzQ

55

3 Comments

Grateful for this conversation and the opportunity to share how we think about real-time systems at scale. At Purplle, performance is not a technical metric — it’s a customer promise. When millions of users expect instant, personalised experiences, every millisecond matters. In this discussion with Nihal Singh and Sonia Benmouffok, we spoke about how our teams use Redis to power some of the most performance-critical workflows — from dynamic pricing and inventory to session management and personalisation — while staying reliable during peak traffic. What stood out for me is not the technology itself, but the engineering mindset behind it: - Designing for scale early - Treating latency as a product feature - Building systems that allow teams to move fast without compromising stability Credit goes to the incredible engineering, data, and platform teams at Purplle who make this possible every day. If real-time performance and personalisation are part of your roadmap, this is a thoughtful watch. Thanks to ETCIO Brand Connect Initiative for facilitating the conversation. #Leadership #EngineeringAtScale #RealTimeSystems #Personalisation #RetailTech #Redis #India

26

I’m honestly amused, by how often people are surprised when I tell them that I use a lot of custom GPTs in my day-to-day work. They are trained on my own posts, video scripts, my technical blogs etc. And then there are a bunch of out-of-the-box GPTs from the GPT Store that I reach for regularly. None of this feels fancy to me anymore. It’s just how I get work done faster, stay consistent, and avoid starting from a blank page every time. What I’ve realized though is that most people either don’t know you can do this, or they don’t quite know how to set it up in a way that’s actually useful. The idea of custom GPTs sounds powerful, but the practical “how” is missing for a lot of folks. So I sat down and put together a very detailed, step-by-step guide. It walks through how I build my own custom GPTs and how I decide which out-of-the-box GPTs are actually worth using for different needs. If this is something you’ve been curious about or slightly confused by, it might help: https://lnkd.in/gSCYxVS5 PS: I’ve been writing a lot more long-form pieces on Substack lately. If you’re enjoying this kind of behind-the-scenes thinking, you might like those too.

117

19 Comments

It is amazing how much of the world's product strategies exist to cater to one simple human behavior: we don't like dealing with friction. Pricing models assume customers won’t spend hours comparing options across ten marketplaces. Subscriptions assume most users won’t fight their way through intentionally complex cancellation flows. Auto-renewals assume people won’t track, review, and renegotiate every policy they hold. This has worked for the longest period because time, attention and patience are finite for us humans. But that reality is about to change given AI Agents have no concept of time or friction. Humans might not have the time to price compare across platforms, their AI Assistants do. Humans might not have the patience to go through 10 steps to cancel their subscriptions, their AI Assistants do. Humans might not know enough to negotiate better policy rates, their AI Assistants do. The principles that drove product design, growth strategies and experiences are likely to get challenged and this is already being seen with every single SaaS app in the market today. #CXAI #AIAssistants

34

3 Comments

Quick question: What percentage of your AI initiatives have made it from pilot to production? If the answer is "not enough," you're experiencing what Gartner calls the Context and Governance Gap. Why MCP matters: • Single integration point for AI agents to access enterprise systems, databases, tools • Governed by design: Explicit permissions, audit trails, compliance-ready • Industry consensus: Anthropic, OpenAI, Microsoft, Google, IBM already aligned CloudHedge #CHAI: Context-Aware Modernization Through our BrewBuzz by CHAI™ series, we're exploring these shifts because they're central to how CHAI operates: 🔹 CHAI Universe™: Intelligent AI Harness for air-gapped orchestration 🔹 CHAI DART™: The most App intelligence for continuous modernization 🔹 CHAI Flow™: Multi-agent collaboration for complex enterprise workflows Context is the difference between: ❌ Experimental AI that can't scale ✅ Production-ready AI that transforms operations The question isn't if you'll adopt Agentic AI? It's whether your architecture can support it securely, at enterprise scale. That's what #CHAI solves. Your thoughts? How are you addressing AI context and governance? #AgenticAI #CHAI #BrewBuzz #MCP #EnterpriseModernization #CloudHedge #Legacy2CloudNative2AgenticApp Sameer Karmarkar Chirag Barhate Ameya Varade Matt Jordan @Ameya Sandip Wane Jack Singer Anurag Rao

25

The timeline is long: a 10–15-year horizon to reach meaningful global scale. But the ambition is to create a borderless, programmable economic network where individuals and businesses control their data and assets, while transacting seamlessly across geographies. https://lnkd.in/gHTsBNCk

3

It is hard to understand Fabric's long-term strategy. On one hand, it was smart to build it around PBI, the leading BI platform. On the other hand, the key to Power BI's success in dethroning Tableau was its' extremely affordable licensing. Competitive licensing is not an element that Fabric's always-on pricing model can boast of at current rates, especially when the likes of Databricks, Snowflake, and other rising platforms are around. Perhaps Microsoft will be content with Fabric always being the 3rd/4th best option?

39

32 Comments

🚀 **Latest AI Industry Highlights for Founders** • **Meta & Scale AI Partnership Shows Strain**: Tensions emerging in high-profile AI data labeling collaboration • **NVIDIA's Continued Dominance**: Reports record sales as AI boom drives unprecedented growth in chip sector • **Enterprise AI Challenge**: New startup Maisa AI raises **$25M** to address the **95%** failure rate in enterprise AI implementations • **Innovation in AI Infrastructure**: - Anthropic launches Chrome-based AI agent - Google & Meta partnering with Ambani to build India's AI backbone - Malaysia's SkyeChip unveils country's first edge AI processor • **AI Safety & Ethics**: - OpenAI co-founder advocates for cross-industry safety testing - Meta implements stricter chatbot rules for teen users - Anthropic settles AI book-training lawsuit with authors #AIInnovation #TechFounders #AIStartups

4

Some days, the data aligns. Some days, it pushes back. Working on JioHotstar’s Core Recommendation Surfaces team means operating at that balance — where viewer intent shifts in real time, content moves fast, and systems must stay reliable at massive scale. For Shivang Singhal, Staff Data Scientist, this is the space where discipline meets curiosity. His work on signals, observability, and ML foundations keeps recommendations both relevant and resilient for millions. His career blends engineering rigour with deep research across healthcare, e-commerce, and advertising—from disease prediction to personalised content and large-scale recommendations. It taught him one core principle: Data matters only when it improves real experiences and drives measurable business impact. What keeps him inspired at JioHotstar is the environment — sharp, passionate colleagues solving problems that turn directly into better moments for viewers across India. Every challenge is a chance to learn, every solution, a step toward deeper customer understanding. At the heart of it, Shivang keeps it simple: Great recommendations start with understanding people. Models are just how we deliver that understanding at scale — viewer after viewer. #LifeAtJioHotstar #JioHotstar #TechatJioHotstar

97

6 Comments