GS

@toddsaunders Can you tell one company that did this and won?

I'm convinced that the biggest vertical SaaS companies of the AI era will not be vertical. They will be horizontal AI harnesses that let the customer build the vertical themselves. For my entire career, vertical SaaS meant the software company learned the industry and built product/marketing specific for it. The moat was domain knowledge, and the vendor was the expert, and the customer was the user. The harness era flips it. Inference has disrupted the moat. The customer is the expert... again. The vendor's job is not to know the industry. It's to build the rails the customer assembles their own software on top of. The industry is about to split in two. The companies that own the rails, payments, identity, compliance, data, become infrastructure. The companies that owned only domain knowledge become a feature on someone else's harness.... There is no third outcome. Vertical SaaS was built on the premise that the vendor was smarter than the customer about the customer's own business. The premise was always weirdly insulting and now it is also obsolete.

Identify the user personas that Hotstar added this plan for?

Solution Architects are now being called Agent PMs....interesting

@InvestorOfJAMMU Bigger picture

Dubai Real Estate prices has crashed. Akshat can plan to buy one more property there.

Added 2 new things today 1) A warning sign for stocks where market manipulation may be happening (news or structure) 2) Themes - tons of them as per Codex's thoughts.

Filters and sorting - Codex 5.3 with cursor

Further improved the daily screener with sector and industries, helping to easily capture the trending sectors -> Industries and the stocks in them. screener-crawal on vercel

Let’s talk about OpenAI Frontier - and why it matters more than “AI features” in enterprise software. Most companies today run on Salesforce, ServiceNow, and Workday. All best-in-class tools. All with AI. Yet work still feels… fragmented. Sales doesn’t fully see hiring. HR doesn’t see revenue risk. IT fixes tickets without knowing the business impact. The problem isn’t the tools. It’s that intelligence is trapped inside silos. OpenAI Frontier is not another app. It’s a layer above your apps. A shared brain that understands: • customers • employees • incidents • workflows All Together. Instead of asking one system for answers, you ask the business. Frontier doesn’t just explain what’s wrong. It finds patterns. Connects cause and effect. Takes action across teams. And learns over time. Your systems stay the source of truth. Frontier becomes the source of understanding and execution. This is the shift happening in enterprise software: Systems of Record → already solved Systems of Intelligence → becoming table stakes Systems of Action across systems → the real moat Companies that adopt this think differently. They don’t ask: “Which tool should handle this?” They ask: “How should the business respond - end to end?” That mindset compounds. In the next decade, winning enterprises won’t be defined by: • more dashboards • more automation rules • more AI buttons They’ll be defined by orchestration. This is what the next enterprise stack looks like.

As a product manager, I’m happy not building generalized products in today’s era, at least until a few customer workflows are automated using quick POCs or MVPs. This enables faster value realization for customers and earlier revenue for the business. Generalized products can come later, once there’s clear signal and scale. This is also where the boundaries between Product Managers, Engineers, Designers, and Solution Architects start to blur.

System of record is no longer the moat in SaaS. That era was defined by companies like Salesforce. System of actions is. That’s where Palantir plays. Seat-based pricing? 🚢⬇️ That ship is already sinking. SaaS is doing great if: You cut costs hard (headcount, software, anything which is useless) AI revenue is growing triple-digits, MoM - which means you have adoption of your AI products, or you have already transformed yourself into an AI-first company. New deals are now tied to value delivered upfront.

@_shankarganesh @storylaneio Awesome, congratulations 🙌

stackoverflow.co/data-licensing That was smart.

Bookmyshow vs Makemytrip - Map view

@_shankarganesh @mrgirish Possible to share?

Align with this a lot

Extremely further away from what I understand.

Hold Assist by Apple is such a great feature. 👏

Actually, I think some bit of automated driving is already there for many cars now. Copied from search 👇 Mahindra does not currently offer a fully self-driving car, but its models, like the XUV700 AX7L, include advanced driver-assistance systems (ADAS) that provide level-one autonomous capabilities, such as adaptive cruise control, lane-keeping assist, and automatic emergency braking