Agentic AdTech

Ad tech has historically evolved around two constraints — data access and creative scalability. We believe the playbooks and approaches for both are being rewritten by foundation models. Meanwhile, distribution continues to fracture as the world moves from search to feeds to chats to agents. The next wave of companies will not merely optimize ad buying; they'll reinvent the medium itself — what an "ad" is, how it's created, who it targets, and how ROI is measured.

Next-Generation DSPs & Ad Networks for the Agentic Web

Conversational platforms like ChatGPT, Perplexity, and Claude have unlocked entirely new surface areas for attention that are contextual, intent-rich, and natively interactive. Traditional display units and keyword targeting feel archaic in this environment. We see an opportunity to build LLM-native ad exchanges that can auction conversational real estate — dynamically inserting context-aware messages into prompts, chats, and agent interactions as these ad units come online.

Autonomous Media Buyers + Longtail Personalization

Imagine one-prompt ad buying: describe your audience, and a platform automatically generates and optimizes creative, targeting, and budget allocation across channels. This would blend AI-driven campaign creation, cross-platform optimization, and real-time feedback loops to make every dollar self-improving. A verticalized DSP could be especially compelling in regulated or high-context industries such as healthcare, pharma, or financial services where creative compliance, copy tone, and approval workflows still require human-in-the-loop systems.

1:1 Scaled Website + Landing Page Generation

Advertising shouldn't end when a user clicks. We're interested in platforms that use LLMs + embeddings to generate real-time, personalized landing pages and website experiences for each visitor — fully integrated into ad data, CRM systems, and behavioral context.

Future of Referrals and Peer-Based B2B Buying

As foundation models seek out new sources of trusted, high-signal data, forums like Reddit, niche industry communities, and expert transcripts have become the new gold mines of credibility. Simultaneously, more B2B buyers rely on conferences and peer networks for authentic product discovery and validation. We believe there's room to build a next-generation Gartner that aligns incentives for trust, leverages peer data + AI insights, and creates a source of truth for evaluating and buying B2B products.

AI-First Attribution & ROI Engines

Most multi-touch attribution models are static, opaque, and decay quickly. We see potential for AI-first attribution systems that use LLMs and code generation to continuously update, simulate, and quantify the marginal ROI of every marketing dollar across platforms, finally bringing transparency and adaptability to performance measurement.

Agentic Platforms for UGC-Driven Growth

Brands increasingly rely on user-generated content to build trust and reach new communities, but sourcing, vetting, paying, and managing creators is still a spreadsheet-and-DM nightmare. We're interested in agentic systems that can autonomously handle the UGC workflow: discover creators, score authenticity and brand fit, negotiate and pay out, track performance, and continuously find the next cohort.