Harkirat Singh

Analogies are powerful - credit for trying. Mine cuts the other way. Having spent years shipping production software whilst at Palantir, here's the truth: vibe coding enterprise tools is like swinging a sledgehammer when the work is sculpting marble. Real enterprises are intricate. The way they build and evolve processes, data, governance, workflows - none of that survives prompted, non-deternimism. The platforms that respect that complexity compound. The ones that don't end up as expensive demos. Most people are finally getting it.

$PLTR is the professional house painting company. Anthropic is like painting a house with a toothbrush. Possible but a bad idea. Open AI is like trying to paint a house with a rake while naked. Ugly and pointless.

@rahul_garg Congrats on the new start @rahul_garg! I was amazed by how much you built in 2 days at the Buildcamp in Bangalore. You’re bound to crush it - all the best.

18 years in enterprise software. Co-founded CloudGo, grew it to one of ServiceNow's largest partners in APAC, exited to RGP. CRM has a fragment. ERP has a fragment. ITSM has a fragment. Buying AI inside each one just makes you faster at guessing in isolation. The problem isn't the AI. It's the data model. That's why I've been watching Palantir for years. First heard the name in a movie. Gerard Butler said it, I went and searched. Been reading their documentation, watching every video, running a dev instance for over a year. Waiting to see if the thing I was seeing was real. One year ago I went to a Palantir build camp in Bangalore. Two days, alone, no prior Palantir experience. Built a complex use case end to end with Palantir FDE's. That's when I stopped watching and started planning. Sixteen years in SaaS trained me to fit customers into the box. Palantir flips it. The platform bends to the customer. Not the other way round. Data first. Ontology first. Outcome first. Tech last. This is the next generation of enterprise platform. The hard problems finally have a platform that can handle them. The ones where data, decisions, AI and human judgment meet. Uriah and I are building @vanyarcom . @PalantirTech Foundry and AIP specialists. The next chapter. Vanyar is live.

@arny_trezzi @chadwahl @SwissRe @connor_deeks Sshhh it’s a consultancy

$PLTR x @SwissRe Stargate crosses +10,000 users: That's +70% of SwissRe employees!

Palantir recently became the military's primary software system for "targeting," and increasingly other aspects of the fight. Since the scary "targeting" OS used scary "AI," we were never going to have a sane conversation about how this works. This story is my best shot at a sane conversation about how this works. First of all, "targeting" isn't shooting. "Targeting" is the exhaustive, bureaucratic process by which we plan, execute, and *assess* shooting. How our planes will get the right munitions. Where they’ll refuel. Cloud coverage. The literal illumination of the moon. Matters of physics. Matters of contingency: the AC-130 gunship needs to be in an overwatch position before the ground assault can happen, but the gunship can’t move into place until the air defenses have been taken out. A “scheme of maneuvers” is determined. Only then does the strike proceed. ^That is only a small snippet of the "targeting" process, btw. As of the start of the conflict with Iran this year, "targeting" was still heavily manual. The glue between disparate processes, even if they featured automation, was: PowerPoint, email, chat, and Excel files. Target lists were relayed in spreadsheets. Sequenced maneuvers sat in Gantt charts in PowerPoint. That bottleneck deepened a relationship with Palantir that started decades ago. Now, hundreds of AI agents in the "Maven" software system do stuff like: make sure there's nothing in the military's historical intelligence that might disqualify a target before a strike. Sources believe AI could've prevented the U.S. strike on a school in Iran, which killed more than 100 children, as it was based on "outdated data." Palantir has been called a lot of things. Evil, dystopian. I've been a tech reporter for almost 10 years, and among my travels, nothing is more dystopian than dysfunctional, manual tech. It literally makes people want to die. It literally causes death. The Pentagon has wanted war AI since Vietnam, for good reason. This is the definitive story of how Palantir built it. piratewires.com/p/how-the-mili…

Need to burry someone. Hold my beer.

🚨🚨 MORE GOOD NEWS ALERT! 🚨🚨 Last week, the Government announced that the NHS Federated Data Platform (FDP) is on track to deliver an incredible £5 for every £1 invested, with an estimated £150 million in benefits annually by the end of the decade! This news comes from a dry UK Government accounting report, but it’s truly a testament to how Palantir technology - hailed yesterday as “the single most important enterprise stack in the world” by Jensen Huang - is revolutionising health outcomes and boosting NHS productivity. While big government projects often face intense scrutiny when they go wrong, the good news rarely gets the spotlight. So far, the FDP has already facilitated 75,000 more operations through smarter scheduling, slashed the longest unnecessary hospital stays by 17.2% by expediting discharges, and this is just the beginning. When was the last time a major government IT project truly delivered on its promises? 🤔

Usage-based pricing creates completely lopsided incentive structure. You are paying for ANY work vs. VALUABLE work. It’s like paying for fuel to run a Hummer H1 on a dynamometer. Organisations with high usage commits need to instrument where their compute is going. How much does that data pipeline, model, or app cost and is it actually generating value for the business? If the cost:value is not 1:3 or higher, mark them for deletion and let the teams fight for it. It’s sad the number of times I’ve seen insanely large data platforms used for lab experiments that keep people “busy” (e.g. models that never see prod), or 1000’s of dashboards that no one looks at.

@PLTRs_Palantir Now ask Grok to list companies that are “CIA funded”

Everyone who screams and cries wolf about $PLTR relationship with ICE, I have a question. Why aren't you screaming and protesting $MSFT $BAH $NOW? And many others (some known, some not so known) Heres a few recent contracts (these are all IT related never-mind other areas, such as gear) Booz Allen received a $12M modification. ICE exercises option period 4 to provide data analytics support to the Homeland Security Investigation's Innovation Lab for operations and maintenance and adaptive maintenance of its analytic platform, tools, and infrastructure. Where is the outrage? DEV TECHNOLOGY GROUP received a near $1M modification to an existing $25M contract for development on Microsoft 365 SharePoint. Where is the outrage? TRIBALCO received a $2M payment on a $40M contract for 3,430 $MSI (Motorola) APX next portable radios for personnel to enable uninterrupted, secure communications during operations, supporting enforcement actions, safeguarding national security, and protecting public safety. Where is the outrage? COUNTERTRADE PRODUCTS received a $10M for ServiceNow licenses, ensuring continued access to the platform essential for IT service management (ITSM). Where is the outrage? ANACAPA MICRO PRODUCTS received a $3M contract for $BB (Blackberry) Enterprise Mobility Annual Maintenance, ensuring continued support and updates for the platform allowing DHS to effectively manage and protect its mobile device. Where is the outrage? All of these are multimillion dollar deals and less than a week old... Where is the outage?

🚨BREAKING 🚨Incredibly proud that the news is out that Palantir has signed a strategic partnership with the MoD worth up to £750million. This agreement will represent a step change in the UK’s defence capability and the future of our defence ecosystem. Palantir will provide the MoD with access to the most advanced AI defence technology, incorporating key elements of the Maven Smart System - the US’s flagship AI programme which was recently acquired by NATO. The results will be transformative. Our Armed Forces will be connected to weapons systems that studies have shown will increase targeting efficiency by 100x, meaning that 20 people can now do work it would have once taken 2,000 to do. That’s a game-changer for the UK and will transform our national defences.

@jweir19 @PalantirTech And still waiting to see some real demos from Salesforce vs. cute characters.

I’m more bullish @PalantirTech than I think I ever have been. I personally think Palantir has a platform that is minimum 10x better than the next best “platform” or product. And then to hear Karp repeatedly say the world is moving to an outcomes based model where companies will take a cut of the output they help generate is top level confidence. They know their platform works and they’re betting their future profits on it. Exactly how a business should be run. $pltr

Organisations that don’t fix their QA/change control bottleneck are going to get left behind. In the past, deploying changes to production meant rigorous development, QA reviews, and user acceptance testing across isolated environments. Followed by intricate promotions from staging to production. Fixing errors in production was fraught with risk, necessitating exhaustive planning, QA cycles, and change approval boards. Today, we can create logical release environments - like in Foundry - where deploying a change or executing a rollback is as seamless as merging or reverting a pull request. This evolution not only minimizes downtime but also improves agility, allowing teams to iterate rapidly while maintaining robust version control. The question CIO’s should be asking: do they need to stick to cumbersome legacy processes - that massively constrain engineering velocity - on modern platforms?

80% of new pilot conversations start with use cases fixated on “predicting the future.” Ironically, this stems from outsourcing strategic thinking to data science and analytics teams, who assume the future is computable—and that’s where the value lies. We have internal Palantir-isms for this: “insight fishing”, “working on the wrong problem”. It’s in the FDE DNA to reshape these aggressively. Don’t try to predict the future. Manage the present well first.

RTFM

When electricity was invented, true value exploded in manufacturing giants - not sleepy utility companies. When the internet arrived, fortunes flowed to consumer apps and SaaS disruptors - not network equipment companies. Value accrues not to commodity resource providers, easily replicated, but to innovators who apply it to transform businesses. The pattern will hold with AI.

A lesson I relearn every few months at Palantir: The edge of our team isn't just acing the 0→1 but also mastering the 1→0 transition gracefully. You might lead a large team or P&L one day, then pivot to ground-zero roles the next or yield to a better idea (not yours). Embracing this mission-driven meritocracy and humility to start anew isn't for everyone. But it's a feature, not a bug, compounding learning, resilience, and non-linear growth.

Insightful piece from @rsylvester1 on the determination at the top of UK Government to seize the AI opportunity. A few of observations: 1. Extremely encouraging to see HMRC and DWP at the top of the priority list. There is literally billions to saved here - and those savings can start to be realised very quickly if the will is there. 2. The examples in the piece are really scratching the surface of what is possible. That isn’t to do them down - two of them are Palantir tech (Bedfordshire Police and Chelsea and Westminster NHS Foundation Trust). But there is SO much more to come on this if we collectively grasp this. 3. This tech revolution should end the bizarre anachronism where delivery roles are seen as less prestigious than policy roles within the UK civil service. As the piece notes, a growing proportion of the civil service needs to be made up of tech specialists directly impacting better public services. The PM was incredibly energised in the meeting at Palantir’s DC office, which the article mentions. It’s heartening to see that appears to be translating into meaningful action on the ground.