TickerTain
TickerTain
NewsroomShortsPortfolioConvergence
NewsroomShortsPortfolioConvergence
←
▶ 34:58 · AI Economics & Business Models · Agent traffic breaks ad-based web economics; Shapley-value attribution enables scalable creator compensation
episode briefing
Sequoia Capital

Parallel’s Parag Agrawal: Building a New Web for AI Agents

2026-08-25 · 3 company · 9 thematic
sentiment
3 bull0 bear0 neu
speakers
parag agrawal

Former Twitter engineer, CTO, and CEO; now building Parallel — web infrastructure for AI agents to keep the open web open. Spent over a decade at Twitter scaling systems and ML infrastructure.

episode shorts · 4

If AI Agents Can Do It Today, What's Left For Tomorrow? | Parag…

Every Page You Publish Now Has Two Audiences | Parag Agrawal, P…

Who Gets Paid When An AI Answers Your Question? | Parag Agrawal…

Agents Will Use The Web 1,000x More Than We Do | Parag Agrawal,…

now playing · AI Economics & Business Models
AI Agentstailwindscore 9/10parag agrawal
AI agents will drive 1000x search volume growth, requiring new search infrastructure
Agents will perform thousands of searches per human prompt, shifting web traffic from human-driven to agent-driven. This requires purpose-built search infrastructure optimizing for quality,…
AI Infrastructuretailwindscore 8/10parag agrawal
Agentic search requires new indexing, ranking, and compute allocation across memory hierarchy
Search for agents demands sub-200ms latency with high-quality token selection from trillion-page corpus, requiring novel model distillation, memory hierarchy optimization, and intelligent c…
Search & Discoverytailwindscore 8/10parag agrawal
Agentic search shifts from keyword matching to structured query understanding and authoritative extraction
Agents send longer, better-specified queries with fewer typos, enabling search engines to retrieve exact authoritative passages (e.g., SEC filing revenue numbers) rather than SEO-optimized…
AI Infrastructuretailwindscore 8/10parag agrawal
Search infrastructure becomes a latency-constrained compute allocation problem across model, agent, and search layers
Parallel optimizes quality, cost, and latency by allocating compute across retrieval, ranking, and model inference layers; new Turbo product achieves 200ms latency for high-quality agentic…
Enterprise AI Adoptiontailwindscore 7/10parag agrawal
Cloud providers will partner with specialized search infra rather than build in-house for agent grounding
Model companies face a build-vs-buy decision for agent search infrastructure; Google Cloud's partnership with Parallel suggests hyperscalers may prefer integrating best-of-breed search prov…
AI Agentstailwindscore 9/10parag agrawal
AI agents will drive 1000x more search queries than humans, requiring new search infrastructure
Agents already perform 5-20 searches per simple query and hundreds to thousands for deep research; background agents monitoring portfolios or prepping meetings multiply search volume by 100…
AI Economics & Business Modelstailwindscore 9/10parag agrawal
Agent traffic breaks ad-based web economics; Shapley-value attribution enables scalable creator compensation
Human attention scarcity underpins ad monetization; when agents replace human visits, content owners lose revenue. Parallel proposes Shapley-value-based attribution to pay content owners pr…
AI Economics & Business Modelstailwindscore 9/10parag agrawal
Shapley value-based micropayments can align incentives between content creators and AI agents
Current ad-based web monetization breaks when agents replace human eyeballs. Parallel proposes using Shapley values to attribute incremental value of each content source to agent outputs, e…
AI Applicationstailwindscore 8/10parag agrawal
Background agents will shift web from pull to push, creating continuous compute allocation on web changes
The next evolution moves from agents pulling search on demand to push-based triggers where agents monitor web changes (satellite imagery, customer commentary, etc.) and automatically initia…