Calcus Cal

Hello $AKAM

$WDAY and Anthropic Announced a New Partnership to Connect Reasoning to the Rules That Run the Business, Wherever People Choose to Work Workday's Sana AI agents now connect with Claude for HR and finance work. People Can Find Workday Information, Take Action, and Approve Work Through Sana Agents in Claude – Workday's Permissions and Business Rules Apply to Every Request, and the User Confirms Every Change. Sana Handles Complex, Long-Running HR and Finance Work and Calls on Claude's Reasoning for Individual Steps, While Keeping the Data Inside Workday

$NVT and $AMD Collaborate on Cooling for AMD Helios Rackscale Solution. nVent will add liquid and air cooling solution to AMD rack-scale Al platform to create a fully integrated system for large Al workloads. The companies will unveil a first look of the open-source solution at the OCP Global Summit. "This collaboration brings nVent's cooling expertise to one of the leading Al platforms in the industry," said Eric Osborn, vice president and general manager of Data Solutions, Cooling & Power at nVent. "We are designing for the platforms our customers are building on, and working closely with AMD is how we make sure the fit right."

Part of me owns $POWL just cause I miss the OG Jerome P

$EXLS - EXLService Holdings, Inc. operates at the intersection of proprietary enterprise data, domain expertise, and AI implementation. Its concentration in insurance, healthcare, banking, and other data-intensive industries is important because successful AI deployment often depends on integrating fragmented operational datasets rather than simply providing a general-purpose model. Their acquisition of iMerit adds capabilities in data annotation, model training, and human-in-the-loop AI systems. That gives EXL additional exposure to the infrastructure required to make enterprise AI accurate and production-ready. Q2 2026 revenue rose 15.6% to $594.8 million, while adjusted EPS increased 22.3%. Management raised expected full-year organic revenue growth to 13–14%, from 10–12%. The revised guidance also includes expected revenue from iMerit. Crucially, these are company-wide growth rates rather than projections for a small experimental AI division. However, GAAP operating margin declined to 14.7% from 15.8%, so the evidence for margin expansion is weaker than the evidence for revenue and adjusted earnings growth. They’re trading at roughly 2.50× EV/sales and 14.9× forward earnings, EXLS looks reasonably valued for a company reporting mid-teens revenue growth and faster adjusted EPS growth. Its forward earnings multiple is also substantially lower than its historical valuation in several recent years. Continued 13–15% organic growth, successful integration of iMerit, and GAAP operating margins stabilizing while revenue scales. Primary risk: AI automation could eventually reduce demand for some of EXL's labor-intensive managed services. The company must capture enough higher-value AI work to offset that pressure.

I think with any infrastructure boom/bubble we’ve seen in the past, a common thread is people buying picks & shovels because those guys experience a huge margin, TAM and revenue inflection. However, like most bubbles eventually margins grow out of hand or demand no longer outpaces supply or innovations are made to optimize and streamline processes, popping the bubble. In the end the people buying and using those picks and shovels end up being the long term winners. I think with AI those clear winners will be $AMZN, $GOOGL, and $META (at least until we see more companies arise that sell new applications for AI). I think for those three $META has the best valuation case and their execution with Muse and public response to it seems outstanding. I now have a good cost basis advantage on $META so going forward this will be a core position for me.

$MSFT is also pushing the gas on AI agents. Microsoft’s Hybrid Intelligence integrates Copilot more deeply into Windows by letting it access local files, documents, and context on the PC, with explicit user permission, to handle tasks and switch to on-device models for faster, cheaper processing.

Deutsche Bank’s Muse model for $META is more interesting for what it leaves out than what it includes. In the optimistic case, Muse reaches ~8% of $META 2030 revenue (~$36B), with Agentic commerce fees driving roughly 3/4s of the upside. The model assumes only low single digit millions of paying users out of $META U.S. Family of Apps base and per-user annual GMV in the $6K–$13K range, with Muse capturing 8–18% of the broader agentic commerce market at 2–5% take rates. It gives no credit for Muse completing purchases through Shopify and Stripe, or for the agent helping businesses generate and spend more on Meta ads. That second loop is the real optionality: Muse can take a cut of the transaction while also feeding incremental dollars back into Meta’s core advertising business. The base case treats Muse as a modest new revenue stream. The unmodeled pieces treat it as a distribution and commerce layer sitting on top of one of the largest consumer graphs in the world. Early download traction is strong, but the revenue math still depends on whether users actually let an agent spend for them at scale and whether Meta captures the checkout rather than just the ad on the way in.

The fact that I’m not even surprised

Mom they’re talking about my stock on TV!! $AKAM

Nice daily structure on $CBRS

10-year Bond yields made a new local high on PMI data 😭😭

Cantor Fitzgerald analyst Andres Sheppard reiterated an Overweight rating and $12.00 price target on $AUR "As of 6/30/26, AUR has completed >6M autonomous cumulative miles (with supervision and driverless combined), and ~440,000 fully driverless miles (vs. >370,000 as of 1Q) after entering commercialization in 4/2025." "More importantly, AUR recently launched its second-generation driverless trucks, which the company is planning to deploy across its commercial network, and for which $AUR's manufacturing partner (Roush) is targeting an annual production run-rate of 1,000 trucks in October. On its 2Q call, AUR disclosed its plan to deploy 20-25 Gen 2 trucks by the end of 3Q, and the company is targeting >200 driverless (Gen 2) trucks in operation by year-end. More importantly, these new kits are expected to result in "a >50% reduction in hardware costs," which we find very encouraging."

Clown world was a big meme and people used to throw that out for almost anything. I genuinely believe we’re living in a 🤡 world today.