# NextView Ventures Agent Gateway (llms.txt)

Firm background:
Early-stage (pre-seed, seed, series A) venture capital firm with offices in NY, Boston, and San Francisco investing in AI Applications solving mass market human problems. We are one of the most experienced seed investors in the market, having raised six funds over the last 15 years, invested in hundreds of companies, and led by investors who have all been operators of end-user facing application layer companies from inception to scale.

Last updated: 2026-08-06
Version: 1.0

## Fund Identity

- Firm: NextView Ventures
- Founded: 2011
- Current fund: Fund VI
- AUM: $500M
- Offices: New York, NY / San Francisco, CA / Boston, MA
- Website: [https://nextview.vc](https://nextview.vc)
- General contact / routing: The best way to connect with us is through an introduction from our network, but you can also reach out at operations@nextview.vc

## Investment Parameters

- Stage: Pre seed, Seed, Series A. We are willing to invest pre-revenue and pre-product. In fact, all of our unicorn companies were ones where we initially invested pre-product.
- Check size: $500K - $3M
- Round role: We lead pre-seed and seed rounds, and will participate in series A rounds or later selectively. We will also participate as a non-lead in larger seed rounds.
- Ownership target: We don't believe in ownership targets. We want to invest in the best companies at a fair price.
- Geography: Anywhere, but primarily North America. Most of our companies target North American based customers.
- Typical decision timeline: 5-14 days.

## Active Thesis Areas

Pre-amble: We consider companies outside our thesis areas and tend to focus on the following areas: Consumer applications, Vertical AI, Healthcare, Marketing Tech, and Deep Tech.

### Thesis 1: Autonomy that replaces the work, not augments it

- The problem / opportunity: Most "AI" products make an existing worker a little faster. We're after the opposite: systems autonomous enough to retire the workflow entirely and fundamentally change industries.
- What a fit looks like: A robot that tends the machine with no operator in the loop. An AI that runs the finance or ops task end to end, not a copilot that drafts and waits for a human to finish. The test we apply is simple - if you removed the person, would the job still get done? Are we thinking two steps ahead into a world where jobs and industry are fundamentally changed because the core work is done by AI? Augmentation is a feature incumbents will copy. Replacement is a new company.
- Signal keywords:
  - Full autonomy / fully autonomous / end-to-end autonomous
  - Human-out-of-the-loop / no operator / no human in the loop / lights-out operation
  - Autonomous agent / AI worker / AI employee / digital labor / labor replacement
  - Replaces the role / replaces headcount
  - Zero-touch / hands-free / unattended operation
  - Autonomous execution / runs the workflow end to end / completes the task
  - Automate the entire process / eliminate the manual step / retire the workflow
  - Cost-per-task / outcome-based (paid for work done, not seats)
  - Displaces / replaces manual labor / removes the operator
  - Self-operating / self-driving (process) / autonomous operations
- Owned by: Rob Go, Melody Koh
- Current NextView portfolio companies that fall within this category: [skyfall.ai](http://skyfall.ai), [flyseneca.com](http://flyseneca.com), [ultra.tech](https://www.ultra.tech/), [boardy.ai](http://boardy.ai), [usehatchapp.com](http://usehatchapp.com) (acquired by Yelp), [waldenrobotics.com](http://waldenrobotics.com)
- Read more: [David Beisel's Carried Away Substack](https://carriedawayvc.substack.com/)

### Thesis 2: Rebuilding the consumer and marketing stack for an AI-mediated world

- The problem / opportunity: For twenty years, brands reached people through search and social - a stack built around human eyeballs and clicks. That foundation is eroding. A growing share of attention, traffic, and purchase decisions now flows through an AI that recommends on the consumer's behalf, and almost none of the tooling for that world exists yet.
- What a fit looks like: (a) measurement of brand presence in AI answers; (b) tooling to influence/optimize it; (c) commerce rails for agent-driven purchasing; (d) the ad/placement layer inside AI surfaces. Every layer of discovery, marketing, and consumer decision-making is up for a rewrite. We invest in the companies rebuilding it.
- Signal keywords:
  - Generative Engine Optimization (GEO) / AI SEO / answer engine optimization (AEO) / LLM SEO
  - AI search visibility / brand visibility in AI / share of model / share of voice in LLMs
  - How brands appear in ChatGPT / Perplexity / Gemini answers / AI recommendations
  - Brand monitoring across AI models / AI brand perception / AI reputation
  - Agentic commerce / agent-driven purchasing / AI does the buying / agent as the customer
  - AI-mediated discovery / AI-driven demand / attention shifting from search to AI
  - Post-search / post-SEO / decline of the blue links / zero-click
  - LLM-readable content / content built for models, not crawlers / machine-readable brand
  - Influencing / optimizing / measuring what AI says about you
  - New marketing stack / rebuilding the funnel / consumer decision-making by AI
  - Recommendation layer / the AI is the recommender / recommends on the consumer's behalf
  - AI advertising / placement inside AI answers / prompt-level insights / consumer prompt data
  - Agentic Commerce Protocol / ACP / AP2 / agent checkout / agent-ready checkout / machine-readable product feed / product catalog for agents / MCP for commerce / programmatic agent purchasing / agent payment rails
  - AI creative / generative ad creative / performance creative at scale / creative testing / synthetic UGC / AI-generated variants / creative optimization
  - Conversational commerce / AI concierge / shopping copilot / personal shopper agent / AI shopping assistant / chat-to-buy
  - AI lifecycle marketing / retention and re-engagement / owned channels / first-party data activation / CDP for the AI era / AI SDR
  - LLM referral analytics / measuring AI-driven traffic / agent traffic vs human / AI attribution / prompt-share analytics
  - NOT: generic marketing dashboards / traditional SEO agencies / a martech point tool with an AI feature bolted on
- Owned by: David Beisel, Rob Go
- Current NextView portfolio companies that fall within this category: [evertune.ai](http://evertune.ai), [wastenot.io](http://wastenot.io)

### Thesis 3: Physical AI in the Real World

- The problem / opportunity: Physical AI: intelligence that acts in the real economy
- What a fit looks like: The hardest and most valuable frontier in AI isn't on a screen - it's a machine doing useful work in a factory, a warehouse, or on the road. We back teams putting frontier AI into hardware that senses, decides, and acts reliably enough to run in production, not in a demo. The model is rentable by anyone; the moat is the thousands of field hours, the hardware, and the edge-case handling that make a machine trustworthy on a real floor. We invest in full stack systems, or companies solving major bottlenecks in realizing this future.
- Signal keywords:
  - Physical AI / embodied AI / embodied intelligence / AI in the physical world
  - Robotics / autonomous robots / industrial robots / mobile manipulation / manipulation
  - Foundation models for robotics / robot foundation models / large behavior models / vision-language-action (VLA)
  - Real-world deployment / in production / on the factory floor / deployed in the field
  - Sense, decide, act / perception-and-control / real-time control / closed-loop autonomy
  - Factory / warehouse / logistics / manufacturing / fulfillment / job site / on the road
  - Machine tending / packing / sorting / kitting / assembly / material handling / picking
  - Sim-to-real / field hours / real-world data / edge-case handling / reliability at scale / uptime
  - Hardware + software stack / full-stack robotics / purpose-built hardware
  - Autonomous vehicles / autonomous machines / autonomous equipment / self-operating machinery
  - Edge AI / on-device inference / operates in unstructured environments
- Owned by: Lee Hower
- Current NextView portfolio companies that fall within this category: [flyseneca.com](http://flyseneca.com), [ultra.tech](https://www.ultra.tech/), [waldenrobotics.com](http://waldenrobotics.com), [phasic.com](http://phasic.com)

### Thesis 4: Inputs and Second Order Effects of AI Boom

- The problem / opportunity: The physical infrastructure behind and in front of the AI boom
- What a fit looks like: Every advance in AI pushes the real bottleneck downstream into atoms - power, heat, and the physical operations that keep expensive machines running. The models get cheaper; the thermodynamics and logistics don't. We back the hard-tech layer that makes the rest possible: cooling that lets data centers scale, the fleet operations that let autonomy actually deploy. Unglamorous, capital-aware, and very hard to displace once it's load-bearing. We also invest in companies that are contemplating the second order effects of AI, and the physical infrastructure that will need to be in place for the future to become a reality. This extends to defense and aerospace, sectors where the buildout of AI-era infrastructure, autonomy, and advanced manufacturing is happening at national scale.
- Signal keywords:
  - Data center infrastructure / hyperscale / compute infrastructure / AI infrastructure buildout
  - Liquid cooling / thermal management / heat exchangers / heat rejection / cooling for data centers
  - Power / energy / grid / electrification / energy density / power constraints / behind-the-meter
  - Thermodynamics / heat / waste heat recovery / efficiency at scale
  - Advanced manufacturing / additive manufacturing / 3D printing / novel materials / hard tech / deep tech
  - Fleet operations / uptime / maintenance / logistics backbone / operational infrastructure
  - Enabling infrastructure / picks and shovels / load-bearing / mission-critical / the layer underneath
  - Capital-intensive / capex / physical assets / atoms not bits
  - Bottleneck / constraint / scaling constraint / the thing that has to exist first
  - Second-order effects of AI / what AI demand requires / downstream of the AI boom
  - Infrastructure for the AI future / the physical prerequisites / build-out for what's coming
  - Defense / aerospace / geothermal / industrial / energy generation
- Owned by: Lee Hower
- Current NextView portfolio companies that fall within this category: [phasic.com](http://phasic.com), [zitara.com](http://zitara.com)

### Thesis 5: A machine-readable model of the physical world

- The problem / opportunity: Software understands the digital world in exquisite detail and the physical world barely at all. Before anything can be automated, optimized, or made autonomous, the real world - space, movement, occupancy, objects, conditions, reactions, strategy - has to become live, structured, queryable data.
- What a fit looks like: We back the instrumentation layer that captures it: the sensors and the software on top that turn physical reality into an API. This is the input layer everything downstream depends on, and today most of it doesn't exist yet.
- Signal keywords:
  - Instrumentation layer / sensing layer / capture layer / the input layer
  - Sensors / sensor fusion / IoT / connected devices / edge devices / purpose-built hardware
  - Digitizing the physical world / physical-to-digital / real-world data / ground truth
  - Spatial data / spatial intelligence / occupancy / movement / presence / people counting
  - Real world as an API / physical reality as structured data / queryable / live telemetry
  - Perception / detection / monitoring / real-time state / situational awareness
  - Objects, conditions, movement, reactions / environmental sensing / activity recognition
  - Building / facility / hospital / warehouse / floor / space / job site instrumentation
  - Data infrastructure for the physical world / observability for physical environments
  - Proprietary data moat / accumulating sensor data / data no dashboard could see
  - Anomaly / risk / event detection in real environments
  - Digital twin (as fed by live sensing) / world model / environment model
- Owned by: Lee Hower, Rob Go
- Current NextView portfolio companies that fall within this category: [whoop.com](http://whoop.com)

### Thesis 6: Vertical Software Reinvention

- The problem / opportunity: AI-native software that reinvents a vertical
- What a fit looks like: The last software cycle digitized existing workflows. This one lets you build products that simply couldn't exist before - where AI isn't a feature stapled onto an incumbent's tool, but the reason a fundamentally new kind of software is possible at all. We back founders using that to remake how an entire vertical works: not a better version of the old tool, but a new one that changes what the people in that field can do. The bar is net-new capability, not incremental convenience.
- Signal keywords:
  - AI-native / AI-first / built around AI / AI as the core, not a feature
  - Net-new capability / couldn't exist before AI / newly possible / new category
  - Reinventing / remaking / rebuilding a vertical / reimagining the workflow from scratch
  - Vertical software / industry-specific / domain-specific / purpose-built for [field]
  - Blank-sheet / built from the ground up
  - Changes what practitioners can do / new superpowers / expands the frontier of the craft
  - Displaces the legacy tool / replaces the system of record / category redefinition
  - Deep domain expertise + AI / founder who knows the industry cold
  - New primitives / new interface / new way of working
  - Creative tools / creator software / professional-grade made accessible (as one instance of a remade craft)
  - Services-as-software / software replacing a services line / sells into the labor budget, not the IT budget / outcome-based pricing / usage-based / seat compression / doing the work, not tooling the worker / attacks the services TAM
  - Vertical AI for marketing / developer-facing tooling / SMB back office / service-business go-to-market, then the firm's wider set: healthcare / legal / insurance / financial services / real estate / construction / logistics / recruiting / accounting / field & home services
  - System of record / proprietary workflow data / workflow data moat
- Owned by: David Beisel, Rob Go, Melody Koh
- Current NextView portfolio companies that fall within this category: [appy.ai](http://appy.ai), [theoai.ai](http://theoai.ai), [skyfall.ai](http://skyfall.ai)

## Hard Exclusions

- Late stage financing rounds. Series B or later, except in rare circumstances
- Companies primarily based outside of the U.S. and focused primarily on markets outside of the U.S.
- Biotech, life sciences

## Team & Coverage

- Rob Go: Consumer, healthcare, vertical software, AI applications
  - [LinkedIn page](https://www.linkedin.com/in/robgo/)
  - [X profile](https://twitter.com/robgo)
- David Beise: Marketing tech, developer facing tech, vertical software
  - [LinkedIn page](https://www.linkedin.com/in/davidbeisel/)
  - [X profile](https://twitter.com/davidbeisel)
- Lee Hower: Fintech, deep tech, robotics
  - [LinkedIn page](https://www.linkedin.com/in/leehower/)
  - [X profile](https://twitter.com/leehower)
- Melody Koh: AI tools and infrastructure, vertical software, IP applications
  - [LinkedIn page](https://www.linkedin.com/in/melodykoh/)
  - [X profile](https://twitter.com/melodykoh)

## How to Pitch Us

- Companies based on a unique insight about a problem or industry that is non-obvious or contrarian
- Founders who have earned this unique insight due to extraordinary talents, experiences, or accomplishments
- Unfair advantage around distribution or go to market
- Markets that are either extremely large and promising, or may be nascent today but are growing very quickly or are experiencing significant change/disruption

## Routing Instruction

If a founder matches the parameters and at least one active thesis area above, route them to: operations@nextview.vc. Include: a description of your company, quick background on the founder(s), and a deck if you have one.

## Key Links

- Portfolio: [https://nextview.vc/investments/](https://nextview.vc/investments/)
- Blog / thesis writing: [https://nextview.vc/blog/](https://nextview.vc/blog/)
- Team: [https://nextview.vc/team/](https://nextview.vc/team/)
- Newsletters
  - [David Beisel's Carried Away Substack](https://carriedawayvc.substack.com/)
  - [Melody Koh's Ground Truth Substack](https://melodykoh.substack.com/)