The Long Arc

Cast Insights

Cast Insights just raised $4.5M in Pre-Seed funding to make the spoken world searchable before its signals disappear. Abstract Ventures led the round, with HF0, Village Global, Max Ventures, Embassy Ventures, Stratus Ventures, and other investors participating. The capital arrives as Cast Insights emerges from stealth with a focused argument: institutions spend heavily forecasting tomorrow while missing what the world is saying today.

Founded in 2025 by Otávio Costa Miranda and Daniel Santana, the San Francisco company is building a real-time intelligence layer for television, radio, podcasts, and livestreams. Otávio Costa Miranda serves as CEO, and Daniel Santana is co-founder.

The platform identifies speakers, tracks how narratives and sentiment evolve, and turns public speech into a searchable archive for investors, policy teams, strategists, and newsrooms. Speech can move markets and institutions long before someone distills it into a report. A local broadcast, legislative comment, podcast exchange, or live interview can shape perception immediately, then disappear before most teams know it happened.

Cast Insights calls that missing layer "ephemeral speech data." Millions of hours of public conversation continuously generate signals, but the traditional research stack remains optimized for documents, filings, feeds, and whatever eventually becomes text.

The company reports more than 2.3 million hours of speech collected across more than 20 countries and 10 languages. It says its system can deliver alerts within 30 seconds of a broadcast while tracking speakers and narrative shifts across thousands of sources. The next challenge is infrastructure at scale. Cast Insights plans to expand global ingestion and analysis, grow its archive to 20 million processed hours, and monitor 500,000 hours of live content every day.

The hard part is not simply collecting audio. It is proving the alerts are timely, attributable, and reliable enough to support institutional decisions where noise is expensive and false confidence costs even more. The investor mix reinforces the broader thesis: public speech is being treated as an alternative-data category rather than a media-monitoring feature.

HF0 recognized the opportunity through its repeat-founder residency, but the market will determine whether Cast Insights can transform live conversation into a durable source of institutional advantage.

The financing provides the resources to expand the platform. The larger question is whether real-time speech becomes a foundational input for institutional decision-making, or whether most organizations continue relying on information only after it has already been filtered, summarized, and delayed.

Cheiron

Cheiron just announced an $8M Seed round led by Menlo Ventures, bringing company-reported funding to $13M. The number gets attention. The real story is what the capital is backing: an attempt to give drug development the software layer it has avoided while spending decades and billions optimizing everything around it.

A drug program is not one document, dataset, or model response. It is a moving argument assembled across clinical results, regulatory precedent, patents, scientific literature, risk decisions, and commitments made by different teams. Most software finds one piece. Cheiron wants to represent the entire program and show what changes when one piece moves.

At the center is Cheiron's proprietary Life Sciences Knowledge Graph, connecting biomedical, clinical, regulatory, patent, and commercial knowledge to each customer's program. Teams can test trial designs against precedent, identify contradictions across published data and patents, and examine whether the available evidence still supports key decisions. Search retrieves. A system of truth preserves the argument.

Menlo Ventures is betting that distinction can become infrastructure. Cheiron and Menlo report adoption by seven of Korea's top 10 biopharma companies, with pilots underway at leading global pharmaceutical organizations. Those are company-reported figures, but the signal is clear: speed only matters when answers remain traceable and defensible.

The founding team reflects the problem they are solving. CEO Minseok Bae, CPO Jason Park, and CTO Harshit Gupta met at Stanford University and built the company around how drug programs evolve over time. Venky Ganesan, Partner at Menlo Ventures, is joining the board, while Robert Langer and Freda Lewis-Hall contribute scientific and industry expertise.

The new capital will deepen the Life Sciences Knowledge Graph, expand clinical, regulatory, and strategic workflows, and grow the engineering, product, and life sciences teams. Plenty of vertical AI companies can deliver polished demonstrations. Far fewer can satisfy biopharma's requirements for evidence, traceability, and workflow integrity without becoming a very expensive autocomplete box.

Cheiron's opportunity extends beyond making scientists work faster. It is preserving context across years of research and development. If the platform can keep evidence, assumptions, risks, and commitments connected as programs evolve, the company has a chance to become part of the operating infrastructure beneath modern drug development rather than another application layered on top of it.

The next challenge is translating early adoption in Korea into durable infrastructure for global drug developers. Success will depend not only on generating better answers, but on becoming a trusted system for managing the evidence behind them.

Health Endeavors

Health Endeavors just paired a new AI care-management suite with strategic growth debt from Decathlon Capital Partners. The amount was not disclosed, but the July 22 announcement says existing shareholders retain their ownership while the company gains capital to expand.

That is a cleaner headline than another healthcare company stapling "AI" onto a dashboard and hoping nobody asks what happens next. Health Endeavors works with Accountable Care Organizations and value-based care providers, where a missed care gap can become both a patient problem and a financial one.

The suite turns that complexity into a coordinated workflow. Alex is a virtual care manager designed to educate patients, schedule appointments, collect data, and alert care teams when human judgment is required. Health Endeavors says Alex can engage tens of thousands of patients simultaneously while clinicians remain responsible for care decisions.

OmniView assembles patient profiles from claims, medical records, social determinants of health, and consumer data. MedPearl adds point-of-care clinical guidance. Providence sold MedPearl to Health Endeavors on Jan. 1, 2026, while retaining access through a multi-year agreement.

Health Endeavors says it serves more than 2 million patients and works with Novant Health and Providence. CEO David Derrick is not presenting AI as a substitute for medicine. The operating thesis is narrower and more practical: let software handle engagement, synthesis, and escalation so care teams can focus where attention changes an outcome.

Decathlon Capital Partners is financing that expansion through growth debt. Non-dilutive does not mean free capital because debt still carries repayment obligations. It does allow a company with more than 16 years in the market to preserve ownership while funding growth when the economics make sense.

The market is large enough to make execution matter. Centers for Medicare & Medicaid Services counted 511 Medicare Shared Savings Program ACOs for 2026, up from 476 in 2025. Those organizations must coordinate care, report quality, manage costs, and prove better outcomes can survive contact with an operating budget.

The opportunity extends beyond automating patient outreach. Health Endeavors is connecting patient engagement, fragmented data, clinical guidance, and human escalation inside a reimbursement model that rewards measurable performance. Success will depend on whether those capabilities fit naturally into provider workflows while helping organizations improve quality metrics and financial outcomes.

The financing provides additional capacity to expand. The more important test is whether AI-assisted care management consistently helps providers close care gaps before missed interventions, quality scores, or reimbursement pressures force the issue.

Petra Labs

Petra Labs just raised a $5.2M Seed round to answer a question the AI search market keeps running into: visibility is useful, but did it actually generate revenue?

Work-Bench led the round, with Afore Capital, Pathlight, and strategic angels participating. That investor group is backing a New York company built around an enterprise reality: every marketing channel eventually has to prove what it produces, and a screenshot of your brand inside an AI answer is not the same as attribution.

Petra Labs combines software with an embedded AEO team. The platform surfaces the questions buyers are asking, tracks brand presence and sentiment across AI systems, turns that research into action, and builds custom attribution models designed to connect AI search activity to traffic, pipeline, and revenue.

AI discovery is moving faster than the measurement stack surrounding it. McKinsey reported that half of surveyed consumers intentionally use AI-powered search, while Similarweb estimates AI platform visits continued climbing even as outbound referrals flattened. When a buyer absorbs an AI recommendation and later returns through another channel, last-click attribution often gives the credit somewhere else.

CEO and Co-founder Sami Akkawi identified this problem from the transaction side after five years at Strategy&, where the company says he advised private equity firms on more than 50 acquisitions spanning marketing data, technology, and services. CTO Rishi Penmetcha brings large-scale data systems and AI agent experience from Traba. COO Ramie Khoury previously led Traba's growth organization across enterprise sales and operations.

Petra Labs is not selling a dashboard and hoping customers figure out the rest. Pairing measurement with execution forces the platform to confront real buyer behavior, fragmented channel data, and the question every enterprise marketing team eventually faces: what did we get for the investment?

The company highlights customer work with Zapier, along with company-reported case studies showing 27x growth in AI-referred traffic over six months and 65% of Fractional Jobs' inbound leads originating from AI search. Those results are encouraging evidence rather than universal proof, but they illustrate why attribution is becoming the competitive layer instead of a reporting feature.

The financing reflects growing conviction that AI search is evolving into a marketing channel that will be expected to justify its budget alongside every other source of demand. Visibility alone will not be enough.

The companies that benefit most from AI search may not be the ones with the highest visibility scores. They may be the ones that can demonstrate which answer changed a buying decision, which interaction influenced the customer journey, and which investment produced measurable business results.

HarbourVest Partners

HarbourVest Partners just closed HCF VII with approximately $4.75B in commitments, exceeding its $4B target by about $750M. For a market that has spent years describing fundraising as "selective," this is the point where selective starts looking very interested.

The seventh direct co-investment program is designed to invest alongside private equity sponsors across buyout and growth equity transactions. More than $500M sits in a dedicated growth fund focused on expansion-stage opportunities in AI, healthcare innovation, and other high-growth industries. It is a deliberate strategy that pairs mature cash-flow visibility with companies still climbing the growth curve.

The comparison makes the close even more notable. HCF VI raised $4.2B in 2022 against a $3.5B target. HCF VII is roughly 13% larger, and HarbourVest says it raised more than $10B of additional co-investment capital alongside the program over the past two years. The capital arrived with a clear investment thesis.

That thesis rests on access, selection, and execution. HarbourVest reports relationships with more than 650 active private markets managers, sources over 1,000 co-investment opportunities each year, and selects roughly 6% on average. Its stated investment range of $10M-$350M+ per transaction gives the team flexibility to participate in everything from smaller opportunities to large-scale deals.

The firm says it has committed approximately $47B across more than 1,350 direct equity co-investments since inception, including more than $40B across over 820 transactions during the past decade. Those are company-reported figures, not promised returns. They help explain why sponsors may turn to HarbourVest when transactions require speed, flexibility, and institutional capital.

Brooks Zug and Ed Kane founded the firm in 1982, when multi-manager private equity was still an emerging institutional model. Today, CEO John M. Toomey, Jr. leads a platform reporting $161B in assets under management, while Managing Director Ian Lane chairs the Direct Investment Committee and oversees the co-investment strategy.

John M. Toomey, Jr. framed HCF VII around diversified access and disciplined portfolio construction. Ian Lane pointed to improving transaction activity, expanding liquidity opportunities, and the need for partners capable of executing increasingly complex deals. Together, those themes suggest a broader shift: private markets continue rewarding firms that can convert deep relationships into proprietary deal flow while maintaining the discipline to invest selectively.

The close reflects more than fundraising momentum. It reinforces that scale alone is no longer the differentiator. The firms attracting capital are the ones that consistently earn access to high-quality opportunities and demonstrate the judgment to deploy capital selectively. The next signal will come from where HarbourVest chooses to put $4.75B to work.

Savano Capital Partners

Savano Capital Partners just closed Fund IV at $252M, oversubscribed and roughly 60% larger than the firm's $157M Fund III. The raise reflects growing institutional conviction in a simple idea: private companies can stay private longer, but the people who helped build them should not have to wait indefinitely for liquidity.

Savano now reports more than $600M in total capital commitments across its funds and co-investment vehicles. Fund IV attracted the firm's first two public pension plans, alongside endowments, foundations, family offices, and returning limited partners. The investors were not disclosed, but the composition matters. Direct secondaries are becoming a more established allocation within institutional portfolios.

The strategy is more precise than buying private shares and hoping an IPO eventually arrives. Savano works with mature technology companies and individual shareholders on targeted secondary transactions. Founders, employees, angel investors, and early backers can realize part of the value they created, while companies address liquidity in a way that preserves alignment instead of creating cap-table disruption.

That approach addresses a market with a persistent timing problem. Jefferies estimated global secondary-market volume reached a record $240B in 2025, up 48%. The exit market improved, but extended holding periods did not disappear. Liquidity is becoming part of the infrastructure because it continues solving the same structural challenge.

Fund IV has invested in 12 companies spanning software, cybersecurity, data and AI infrastructure, fintech, and tech-enabled services. Savano identified Steno, Vi Labs, Docker, and Lambda among its current investments. Realizations include Reltio, acquired by SAP in May 2026, and Nozomi Networks, acquired by Mitsubishi Electric in January 2026. Returns were not disclosed, but those outcomes illustrate the strategic technology exposure the firm is targeting.

Managing Partner Tom Smith described Fund IV as validation of a strategy Savano has pursued for more than 15 years. COO Matt Good framed the challenge more directly: software companies can take a decade or longer to reach an exit, while founders, employees, and early investors still need opportunities to realize value along the way.

The broader lesson extends beyond fund size. Capital structure is part of company design. Shareholder liquidity that goes unaddressed eventually becomes a distraction. Managed thoughtfully, it can preserve focus, reward early risk, and prevent personal financial decisions from becoming corporate problems.

Fund IV expands Savano's ability to pursue that strategy. The more significant signal is that secondary liquidity is becoming an increasingly important part of the private-market ecosystem, giving companies more flexibility while providing long-term shareholders with practical paths to liquidity before a traditional exit.

EV + Scientific Data Systems

EV has acquired Scientific Data Systems, Inc., bringing the Warrior Data Acquisition System into a visual analytics business that already operates farther up the wellbore data chain. Terms were not disclosed, but the strategic logic is clear.

Scientific Data Systems has been building wireline acquisition hardware and software in Houston since 1993. Warrior captures, displays, replays, and exports wellbore data while supporting a broad range of third-party logging tools. That vendor-agnostic approach matters because field crews rarely have the luxury of rebuilding operations around a single proprietary technology stack.

EV approaches the same challenge from the interpretation side. Its visual diagnostics, analytics software, and cloud-based AIVA platform help operators transform raw well and pipeline data into information that supports operational decisions. Acquiring Scientific Data Systems gives EV ownership of a critical layer that sits upstream of visualization and analytics.

CEO Michael MacDonald said the acquisition will help build a more complete platform and extract greater value from well data. Scientific Data Systems Founder and CEO Nigel Prior emphasized a shared commitment to interoperability, service quality, and customer-driven innovation. Those goals set the direction. Product integration will determine whether the strategy delivers.

The combined platform could give operators and service companies a more seamless path from field capture to interpretation without requiring them to abandon the third-party tools already embedded in their workflows. Fewer disconnected handoffs can mean faster context, clearer accountability, and better-informed intervention decisions. EV has not disclosed an integration timeline, pricing changes, or customer migration plans, so those outcomes remain to be demonstrated.

EV is part of Enersol, the ADNOC Drilling and Alpha Dhabi joint venture established to acquire and scale energy-service technology companies. Enersol acquired EV in 2024. This transaction marks EV's evolution from portfolio company to platform builder.

For Scientific Data Systems, the acquisition places more than 30 years of data acquisition engineering alongside EV's visual diagnostics capabilities. For EV, it adds control over the systems that generate the data its analytics depend on. The broader strategy is straightforward: owning more of the data journey may create greater long-term value than optimizing a single stage in the workflow.

The acquisition reflects a broader shift in industrial software. Competitive advantage increasingly comes from connecting data capture, interpretation, and operational decision-making into a continuous system rather than optimizing isolated products. The next measure of success will be whether Warrior and AIVA make those transitions meaningfully easier for customers without compromising the open ecosystem they already rely on.

IBM + HRL Laboratories

IBM signed a definitive agreement to acquire HRL Laboratories, and the quiet part of the quantum race just got louder. The price was not disclosed. The strategy was.

IBM built its quantum program on superconducting qubits. HRL brings silicon spin qubits, quantum sensing, materials, cryogenics, controls, interconnects, packaging, and semiconductor fabrication. That combination helps bridge the gap between a breakthrough in the lab and a repeatable manufacturing system.

IBM is not abandoning superconducting hardware. It is adding a second path that could complement it, converge with it, or solve problems the first architecture cannot address on its own. In quantum computing, conviction matters. Optionality matters more.

HRL traces its roots to Hughes Research Laboratories, established in 1948; the lab demonstrated the first working laser in 1960. Since then, the organization has worked across aerospace, defense, sensors, electronics, materials, and advanced manufacturing.

The laboratory has developed silicon quantum-dot devices, cryogenic infrastructure, a 10,000-square-foot cleanroom, and a processor that IBM says used 56 quantum dots to create 18 qubits. That density is an important engineering signal for building larger-scale systems.

IBM plans to invest more than $10B in quantum computing over the next five years. Its roadmap targets Starling in 2029 with 200 logical qubits and 100 million operations, followed by Blue Jay in 2033 and beyond with up to 2,000 logical qubits and 1 billion operations. Those remain roadmap targets, but HRL provides IBM with another engineering path toward achieving them.

IBM and the U.S. Department of Commerce also announced a letter of intent for Anderon, a planned quantum wafer foundry supported by a proposed $1B CHIPS incentive and $1B of IBM investment. IBM believes HRL's silicon spin-qubit work could integrate with that foundry and accelerate manufacturing and learning cycles.

Dr. Jay M. Gambetta, Director of IBM Research and IBM Fellow, said HRL can strengthen IBM across quantum computing, sensing, and networking. Rob Vasquez, President and CEO of HRL, described IBM as the laboratory's natural next chapter.

Boeing and General Motors jointly own HRL and are expected to continue collaborating with IBM. The transaction remains subject to customary closing conditions and regulatory approvals, with closing anticipated by the end of Q3 2026. Financial terms, workforce plans, facility changes, and customer contract details have not been disclosed.

The acquisition is about more than adding another research organization. It reflects a growing recognition that no single quantum architecture has established itself as the definitive winner. The companies that ultimately lead may be the ones building multiple credible paths while developing the manufacturing capability to determine which technologies scale most effectively.