Teahose.
SIGN IN
NEW HERE — WHAT TEAHOSE DOES
We read the entire AI & tech firehose — so you don't have to.
PODPodcastsAll-In, No Priors, Acquired…
NEWNewslettersStratechery, Newcomer…
PAPPapersPhysical AI research
PHProduct Huntdaily launches
VCInvestor ScoutSequoia, a16z, Benchmark…
CLAUDE DISTILLS →
7 reads, 30 sec each — free, 6 AM ET.
+ a live graph of the companies, people & themes underneath.
HOME/POSTROUND/Series A activity: Week of Octob…
NEWS
// NEWSLETTER ISSUE
POSTROUND

Series A activity: Week of October 18, 2026

DATE October 26, 2026SOURCE POSTROUNDPARTICIPANTS POSTROUND
// SUMMARY

1. Key Themes

Theme: Mega-rounds concentrated in AI infrastructure and compute efficiency

A few enormous Series As dominate the weekly total, skewing the average far above the median and showing capital concentrating in AI efficiency and foundational layers.

  • "There were 31 Series As in the last week that raised a total of $2.3 billion. The average deal size was $73.4 million while the median was $20.7 million."
  • "Flapping Airplanes raised $1 billion at a $6 billion valuation from Index Ventures, Lux Capital on October 21, 2026. AI research lab building GPU megakernels and training methods to make large models more efficient."
  • "DoubleAI raised $200 million from Bessemer Venture Partners, Lightspeed Venture Partners on October 19, 2026. AI code-execution models that optimize how software runs on CPUs and GPUs."
  • "Bessemer and Lightspeed co-leading a $200M Series A into Amnon Shashua's newest company is a bet that AI can sit between code and silicon, not just write software."

Theme: AI + physical world (manufacturing, robotics, lab automation)

Investors are funding AI that attacks physical bottlenecks, not just digital ones.

  • "Lightspeed leading a $215M Series A at an $865M valuation into an AI-and-robotics chemistry CRO is one of the clearest bets that drug discovery bottlenecks are physical, not just computational."
  • "Accel, Altimeter, and Khosla co-leading a $160M Series A at a $1.5B valuation into Bob McGrew's manufacturing AI company is a bet that factory changeover can move at software speed."
  • "Eka Robotics raised $150 million at a $1.4 billion valuation from Andreessen Horowitz, F-Prime Capital on October 20, 2026. Vision-force-action robotics for general-purpose physical manipulation beyond human performance."
  • "Mach9 raised $34 million from Bessemer Venture Partners, First Round on October 21, 2026. Geospatial AI production software for faster civil engineering models."

Theme: AI security and enterprise AI risk tooling

Multiple mid-sized rounds target securing AI deployments and automating security operations.

  • "Opsin raised $21 million from Race Capital, SYN Ventures on October 19, 2026. Enterprise AI security and risk visibility platform."
  • "Mave raised $28 million from Red Dot Capital Partners on October 22, 2026. AI-powered security operations automation platform."
  • "Ecomtent raised $11.2 million from Puma Growth Partners on October 21, 2026. Generative Engine Optimization software for enterprises and AI marketing."

Theme: Defense, resilience, and "bring manufacturing back" tailwinds

Defense-adjacent and reindustrialization wedges are attracting capital, with Israel notably active.

  • "The defense and data-center supply-chain wedge, plus a Founders Fund / Accel seed history, makes this one of the clearest bring-manufacturing-back AI applications of the week."
  • "High Lander raised $16 million from EDGE Group on October 19, 2026. Drone operations software for fleet management and traffic control."
  • "Arda... already deployed inside Anduril's ArsenalOS, with a roadmap toward AI-programmed robotic lines."

Theme: Fintech and payments infrastructure continues to attract steady, smaller checks

Embedded finance and stablecoin-adjacent rails remain a consistent category.

  • "Mural raised $15 million from Tether on October 20, 2026. Instant global payments platform leveraging blockchain technology."
  • "CapitalOS raised $21.9 million from LGVP on October 22, 2026. Embedded spend management infrastructure for B2B platforms."
  • "Prism Data raised $20 million from Story Ventures on October 19, 2026. Cash flow underwriting infrastructure and risk scoring analytics platform."

2. Contrarian Perspectives

Series A is no longer a small-check stage

The headline stats show Series A deal sizes diverging sharply from the "typical" round, which challenges the conventional notion of a Series A.

  • Evidence: "The average deal size was $73.4 million while the median was $20.7 million." Flapping Airplanes alone ($1B at $6B) accounts for ~43% of the $2.3B total, and three rounds above $150M account for well over half.
  • Implication: Founders benchmarking against the average will be misled; the median ($20.7M) is the more realistic anchor, while the tail reflects a separate, lab-style capital regime.

Bottlenecks in AI-for-science may be physical, not computational

Counter to the "more compute and models" narrative, the newsletter frames wet-lab throughput as the constraint.

  • "one of the clearest bets that drug discovery bottlenecks are physical, not just computational."
  • "Warp makes and tests molecules for pharma and biotech customers rather than running its own pipeline, so the data loop compounds while customers keep their programs."
  • Reasoning: A service/CRO model, usually viewed as low-multiple, is valued at $865M because it captures proprietary experimental data while avoiding pipeline risk.

Efficiency over scale in frontier AI

Rather than funding bigger models, the largest round of the week backs making existing compute work harder.

  • "AI research lab building GPU megakernels and training methods to make large models more efficient." (Flapping Airplanes, $1B at $6B)
  • "the week's strongest 'AI that makes existing compute work harder' platform raise." (DoubleAI)
  • Reasoning: Two of the biggest rounds target optimization layers, suggesting investors see software-level efficiency gains as a defensible alternative to the pure scaling race.

3. Companies Identified

DoubleAI

  • Description: AI code-execution models that optimize how software runs on CPUs and GPUs (Ramat Gan, Israel).
  • Why mentioned: One of the "Three Most Interesting Deals"; $200M Series A.
  • Quotes: "DoubleAI's Code Execution Models optimize how programs run on CPUs and GPUs without changing source, with Nvidia and Dell also in the round."

Warp Sciences

  • Description: AI and robotics chemistry CRO making and testing molecules for biotech and pharma (South San Francisco).
  • Why mentioned: $215M at $865M valuation; AI-for-science platform.
  • Quotes: "The OpenAI/Meta/White House founding stack, plus Premji, Mayfield, IVP, Lux, and Menlo in the round, makes this the week's strongest AI-for-science platform raise."

Arda Industries

  • Description: AI software that turns product designs into assembly lines and keeps them current as designs change (Redwood City, CA).
  • Why mentioned: $160M at $1.5B valuation; manufacturing AI.
  • Quotes: "Arda turns CAD and design artifacts into assembly instructions and line plans, already deployed inside Anduril's ArsenalOS, with a roadmap toward AI-programmed robotic lines."

Flapping Airplanes

  • Description: AI research lab building GPU megakernels and training methods for model efficiency (San Francisco).
  • Why mentioned: Largest round of the week ($1B at $6B).
  • Quotes: "AI research lab building GPU megakernels and training methods to make large models more efficient."

Eka Robotics

  • Description: Vision-force-action robotics for general-purpose manipulation (Cambridge, MA).
  • Why mentioned: $150M at $1.4B valuation from a16z and F-Prime.
  • Quotes: "Vision-force-action robotics for general-purpose physical manipulation beyond human performance."

Coherence Neuro

  • Description: Implantable bioelectronic interfaces for recording and treating cancer (Cambridge, UK).
  • Why mentioned: Notable backers including OpenAI and Founders Fund.
  • Quotes: "Implantable bioelectronic interfaces that record and treat cancer with adaptive electrical stimulation."

Mach9

  • Description: Geospatial AI production software for civil engineering (Pittsburgh).
  • Why mentioned: $34M from Bessemer and First Round.
  • Quotes: "Geospatial AI production software for faster civil engineering models."

Opsin

  • Description: Enterprise AI security and risk visibility platform (San Jose).
  • Why mentioned: AI security theme.
  • Quotes: "Enterprise AI security and risk visibility platform."

Mave

  • Description: AI-powered security operations automation (Tel Aviv).
  • Why mentioned: AI security theme; Israeli activity.
  • Quotes: "AI-powered security operations automation platform."

Mural

  • Description: Instant global payments on blockchain (New York).
  • Why mentioned: Backed by Tether.
  • Quotes: "Instant global payments platform leveraging blockchain technology."

CapitalOS

  • Description: Embedded spend management infrastructure for B2B platforms (Belmont, CA).
  • Why mentioned: Embedded fintech.
  • Quotes: "Embedded spend management infrastructure for B2B platforms."

Prism Data

  • Description: Cash flow underwriting infrastructure and risk scoring (New York).
  • Why mentioned: Fintech data infrastructure.
  • Quotes: "Cash flow underwriting infrastructure and risk scoring analytics platform."

Nirva

  • Description: AI wearable pendant for personal journaling and guidance (San Francisco).
  • Why mentioned: Consumer AI hardware with a privacy angle.
  • Quotes: "AI wearable pendant that listens only to the wearer for personal journaling and guidance."

Baba

  • Description: Care coordination pairing older adults with dedicated advocates (New York).
  • Why mentioned: $40M from FirstMark; aging-care services.
  • Quotes: "Care coordination service pairing older adults with dedicated advocates."

Alta

  • Description: AI stylist and personal shopper (Redlands, CA).
  • Why mentioned: Consumer AI commerce.
  • Quotes: "AI stylist and personal shopper generating outfit recommendations."

High Lander

  • Description: Drone operations software for fleet management and traffic control (Tel Aviv).
  • Why mentioned: Defense/drone infrastructure.
  • Quotes: "Drone operations software for fleet management and traffic control."

Quantrol Ox

  • Description: Automated ML control software for quantum technologies (Espoo, Finland).
  • Why mentioned: Quantum tooling layer.
  • Quotes: "Automated machine learning control software for quantum technologies."

Optivolt

  • Description: Shade-tolerant solar technology (San Mateo).
  • Why mentioned: Energy tech raise.
  • Quotes: "Sustainable energy technology with shade-tolerant solar power solutions."

4. People Identified

Amnon Shashua

  • Description: Founder of DoubleAI; known from Mobileye and AI21.
  • Why mentioned: Founder pedigree underpins the $200M round.
  • Quotes: "Bessemer and Lightspeed co-leading a $200M Series A into Amnon Shashua's newest company..." and "Coming out of Mobileye and AI21 DNA, this is the week's strongest 'AI that makes existing compute work harder' platform raise."

Bob McGrew

  • Description: Founder of Arda Industries (manufacturing AI).
  • Why mentioned: Founder-led $1.5B valuation Series A.
  • Quotes: "...into Bob McGrew's manufacturing AI company is a bet that factory changeover can move at software speed."

5. Operating Insights

Position as a data-compounding services layer, not a pipeline owner

Warp Sciences' model shows how service businesses can command premium valuations when they accumulate proprietary data across customers.

  • "Warp makes and tests molecules for pharma and biotech customers rather than running its own pipeline, so the data loop compounds while customers keep their programs."
  • Takeaway: Avoiding customer-channel conflict while retaining a learning loop can unlock both sales velocity and defensibility.

Pick a high-urgency wedge and embed in an anchor customer's stack

Arda's early traction came from being embedded inside a marquee customer's platform in a strategic sector.

  • "already deployed inside Anduril's ArsenalOS" and "The defense and data-center supply-chain wedge."
  • Takeaway: Integration with a category-defining customer and a mission-critical vertical can accelerate credibility ahead of broader horizontal rollout.

Benchmark fundraising against the median, not the average

Founders planning a Series A should calibrate to the typical round rather than headline mega-deals.

  • "The average deal size was $73.4 million while the median was $20.7 million."
  • Takeaway: Most rounds in this set cluster in the $10M-$30M range (e.g., Opsin $21M, Prism Data $20M, Alta $20M); plan and pitch accordingly.

6. Overlooked Insights

Strategic and non-traditional investors are showing up in rounds

Several rounds include corporate or adjacent strategics rather than only classic VCs, hinting at ecosystem alignment.

  • "with Nvidia and Dell also in the round" (DoubleAI); "Coherence Neuro raised $45 million from Founders Fund, OpenAI"; "Mural raised $15 million from Tether."
  • Why it matters: Strategic capital can signal distribution, supply, or technology partnerships, particularly in compute-efficiency and payments rails.

Geographic dispersion: strong activity outside the US hub

A meaningful share of notable rounds originated in Israel and elsewhere internationally.

  • "DoubleAI is based in Ramat Gan, Israel," "Mave is based in Tel Aviv, Israel," "High Lander is based in Tel Aviv, Israel," alongside raises from Cambridge UK, Melbourne, Toronto, Potsdam, London, and Espoo.
  • Why it matters: Israel contributed a $200M mega-round plus two other deals in a single week, reinforcing its role in compute, security, and defense-tech while other international hubs show steady mid-size activity.