Enterprise Weekly #532: Anthropic's Week of Ships, Agentic Infrastructure Content, and More
Work-Bench is a NYC-based $160M enterprise software Seed stage VC fund. If you are or know anyone thinking about founding an enterprise software, SaaS, or infrastructure software startup - we’d love to meet! Please reach out to chat.
Anthropic continues its tear on becoming the predominant enterprise AI lab.
Earlier this week, they unveiled Claude Sonnet 4.5, their most advanced coding model yet. The results speak for themselves: on the SWE-bench Verified benchmark, Sonnet 4.5 achieved an impressive 82% (using parallel test-time compute), cementing its place as best-in-class. On OSWorld, a benchmark designed to measure computer-use capabilities, Sonnet 4.5 scored 61.4%, a huge leap forward from prior Claude models, which had all trailed below 45%.
And the feedback from product leaders in the industry only validates the model’s efficacy in production (both in coding use cases and others):
“For Devin, Claude Sonnet 4.5 increased planning performance by 18% and end-to-end eval scores by 12%— the biggest jump we’ve seen since the release of Claude Sonnet 3.6.” —Scott Wu, CEO of Cognition
“It’s surprisingly efficient at maximizing actions per context window through parallel tool execution, for example running multiple bash commands at once.” — Jeff Wang,, CEO of Windsurf
“Claude Sonnet 4.5’s edit capabilities are exceptional. We went from 9% error rate on Sonnet 4 to 0% on our internal code editing benchmark.” — Michele Catasta, President of Replit
But Sonnet 4.5 wasn’t the only thing Anthropic dropped this week. Anthropic released the Claude Agents SDK, making it easier to build autonomous agents with Claude Code as the blueprint.
We’re undergoing a virtuous cycle of applications and infrastructure, where great applications amass lessons at the frontier and their infrastructure evolves accordingly. As Anthropic productizes that infrastructure like they’re doing with the Claude Agent SDK, opens up vast potential for creativity for the rest of the ecosystem. It will be fascinating to see what developers build with these new tools!
Speaking of AI infrastructure, our teammates Priyanka and Proby just released great pieces of writing on the topic! Priyanka released the September edition of her State of Agentic Systems newsletter, where she dives into how agents are now becoming transactional and durable, enabling real revenue generation. Proby released a blog post where he talked about how applications and infrastructure may be converging in this new agentic era. Give them both a read!
Lastly, we wanted to congratulate our Work-Bench portfolio founder April Koh for being featured on the cover of TIME as one of the world’s most influential rising stars! We invested in Spring Health’s seed round back in 2018 and it has been incredible to watch her lead the company to becoming the most preeminent mental health-care company today.
🗓️Join more Work-Bench events:
🍻Pre-Seed Happy Hour with NYC VCs (October 22nd): Work-Bench is partnering with our friends at Bloomberg Beta, Comma Capital, Opal Ventures, Slow Ventures, and more to bring together the early-stage B2B software community. We hosted the same event last year and are excited to be running it back! This event is for pre-seed & future founders looking to meet cofounders, customers, and collaborators. Sign up here to join us!
📚 Read more news:
WSJ: Walmart Wants More Developers, and More AI Agents to Automate Their Work
TechCrunch: Hacking group claims theft of 1 billion records from Salesforce customer databases
VentureBeat: Google’s Jules coding agent moves beyond chat with new command line and API
Fortune: Spending on AI is increasingly fueled by debt, Goldman Sachs says
CNBC: Meta wants to bring its Business AI chat tools to third-party websites
🧵 Read more threads:
Jon Lehr: Leapfin-Metronome Partnership
Kelley Mak: False Positivity in AI-Generated Code
Proby Shandilya: Probservations S1E3 - Context Engineering
Aaron Levie: AI Agents Transcending Talent Availability Constraints
Shawn Wang: Discussion on Semantic Search
Company: Dialpad
Role: Account Executive, Enterprise
Technology: AI-Powered Customer Communications
Funding: $450M from Work-Bench, Andreessen-Horowitz, ICONIQ, and others
🌟 Company of the Week 🌟
Composite raises $5.6M led by NFDG
Collaboration / Productivity - Seed - San Francisco, CA
Founders: Yang Fan Yun, Charlie Deane
Hupside raises $1.7M led by Ruxton Ventures
Future of Work • Pre-Seed • Washington, DC
Arqh raises $3.8M led by Founderful
Future of Work • Pre-Seed • Zurich, Switzerland
SolidCore.ai raises $4M led by Runtime Ventures
Risk / Security • Seed • Menlo Park, CA
Clarifeye raises $4.7M led by EQT Ventures
Data / AI / ML • Pre-Seed • Paris, France
Tato raises $5M led by Ridge Ventures
Collaboration / Productivity • Seed • Montreal, Canada
Notch.cx raises $7M led by Lightspeed
Future of Work • Seed • Tel Aviv, Israel
DJUST raises $8.2M led by New Enterprise Associates
Sales / Marketing • Series A • Paris, France
Supernova raises $9.2M led by Taiwania Capital
Collaboration / Productivity • Series A • Dover, DE
StackGuardian raises $10M led by TIN Capital
Infrastructure / Dev Tools • Seed & Series A • Mechelen, Belgium
Tie raises $10M led by Innovating Capital
Sales / Marketing • Series A • Miami, FL
Anything raises $11M led by Footwork
Data / AI / Machine Learning • Series A • San Francisco, CA
Gain raises $12M led by The Garage
Collaboration / Productivity • Seed • Tel Aviv, Israel
Datawizz raises $12.5M led by Human Capital
Data / AI / Machine Learning • Seed • San Francisco, CA
GoodFit raises $13M led by Notion Capital
Sales / Marketing • Series A • London, UK
Alex raises $17M led by Peak XV Partners
HR Tech • Series A • San Francisco, CA
Mondoo raises $17.5M led by HV Capital
Risk / Security • Series A • San Francisco, CA
Zania raises $18M led by NEA
Risk / Security • Series A • Palo Alto, CA
Lexroom.ai raises $19M led by Base10 Partners
Future of Work • Series A • Milan, Italy
Polars raises $21M led by Accel
Data / AI / Machine Learning • Series A • Amsterdam, The Netherlands
Paid raises $21.6M led by Lightspeed Venture Partners
Future of Work • Seed • London, UK
Flox raises $25M led by Addition
Infrastructure / Dev Tools • Series B • New York, NY
MAI raises $25M led by Kleiner Perkins
Sales / Marketing • Seed • San Francisco, CA
Moonlake AI raises $28M led by AIX Ventures, Threshold, and NVIDIA Ventures
Data / AI / Machine Learning • Seed • San Francisco, CA
Oneleet raises $33M led by Dawn Capital
Risk / Security • Series A • San Francisco, CA
Dash0 raises $35M led by Accel and Cherry Ventures
Infrastructure / Dev Tools • Series A • New York, NY
Descope raises $35M led by Notable Capital, Lightspeed, Dell Technologies Capital, and Others
Risk / Security • Seed Extension • Los Altos, CA
Axiom Math raises $64M led by B Capital
Future of Work • Seed • San Francisco, CA
Modal Labs raises $87M led by Lux Capital
Infrastructure / Dev Tools • Series B • New York, NY
Eve raises $103M led by Spark Capital
Future of Work • Series B • San Francisco, CA
CyberCube raises $180M led by Spectrum Equity
Risk / Security • Venture • San Francisco, CA
Periodic Labs raises $300M led by a16z
Data / AI / Machine Learning • Seed • San Francisco, CA
Vercel raises $300M led by Accel and GIC
Infrastructure / Dev Tools • Series F • San Francisco, CA
Harness acquires Qwiet AI for an Undisclosed Amount
Risk / Security • Acquisition • San Jose, CA
Insight Enterprises acquires Inspire11 for an Undisclosed Amount
Data / AI / Machine Learning • Acquisition • Chicago, IL
Lightcast acquires Simply for an Undisclosed Amount
HR Tech • Acquisition • London, UK
Vectra AI acquires Netography for an Undisclosed Amount
Risk / Security • Acquisition • Annapolis, MD