Proxis (YC S24

About Proxis (YC S24

Proxis provides an LLM distillation and serving layer optimized for low-latency, on-premise deployment of generative AI. The platform deploys AI email agents that ingest enterprise knowledge bases to auto-draft personalized, on-brand replies, only sending automatically when high confidence thresholds are met.

<problem>Enterprises receive a high volume of emails that require manual drafting, personalization, and follow‑up, consuming significant time for sales, customer‑service, and operations teams. At the same time, deploying large language models (LLMs) for such tasks is expensive, introduces latency, and raises data‑privacy concerns when models are hosted off‑site.</problem> <solution>Proxis delivers a two‑part platform that tackles both challenges. First, it provides a dedicated LLM distillation and serving layer that reduces compute costs and latency while allowing on‑premise or secure cloud deployment. Second, it builds AI email agents that ingest a company’s full email history and connected knowledge bases (CRMs, Notion, Drive, Slack, etc.) to auto‑draft replies that match the user’s tone and style. The agents only auto‑send when confidence thresholds are met, otherwise they remain as drafts for review. By combining model optimization with context‑rich automation, Proxis enables enterprises to handle inbox workloads at scale without sacrificing security or performance.</solution> <features> - LLM distillation and serving optimizations that lower inference cost and improve latency. - AI email agents that ingest full email history and company knowledge to generate on‑brand replies. - Context‑first drafting with live integrations to CRMs (Salesforce, HubSpot), Notion, Google Drive, Calendar, and Slack. - Auto‑send functionality that triggers only when the model’s confidence exceeds a configurable threshold. - Team library for shared templates and context packs, enabling consistent communication across users. - Tiered SaaS pricing (Essential $20 / seat / month) with a free 7‑day trial and higher‑tier plans offering dedicated support and SLOs. - VSCode extension for developers to call optimization functions directly. </features> <target_audience>Proxis targets enterprise sales, customer‑service, and operations teams that rely heavily on email, as well as developers and IT groups needing efficient, secure LLM deployment for internal applications.</target_audience> <revenue_model>Proxis operates a subscription SaaS model with an Essential plan at $20 per seat per month, offering auto‑draft only, and higher‑tier plans that include auto‑send, live context connections, priority support, and dedicated account management.</revenue_model> <traction>Proxis launched in YC Summer 2024 and has been featured in YC Launch posts highlighting its AI email agents and LLM distillation platform. The company lists trusted customers such as Google, TCS, and Walmart on its homepage. As of November 2025, the platform offers a free trial and reports active enterprise adoption across multiple inboxes per sales rep.</traction> <sources> - https://www.ycombinator.com/companies/proxis - https://www.ycombinator.com/launches/OnD-proxis-ai-email-agents-with-your-context - https://www.ycombinator.com/launches/Lkv-proxis-the-first-dedicated-platform-for-llm-distillation-and-serving - https://proxis.ai/ - https://www.proxis.ai/#about-intro - https://www.proxis.ai/#how-it-works - https://www.proxis.ai/#pricing-2 - https://youtu.be/9yamssS2oiA </sources>

What does Proxis (YC S24 do?

Proxis provides an LLM distillation and serving layer optimized for low-latency, on-premise deployment of generative AI. The platform deploys AI email agents that ingest enterprise knowledge bases to auto-draft personalized, on-brand replies, only sending automatically when high confidence thresholds are met.

Where is Proxis (YC S24 located?

Proxis (YC S24 is based in San Francisco, CA,US.

When was Proxis (YC S24 founded?

Proxis (YC S24 was founded in 2024.

How much funding has Proxis (YC S24 raised?

Proxis (YC S24 has raised 500000.

Location
San Francisco, CA,US
Founded
2024
Funding
500000
Employees
3 employees
Major Investors
Y Combinator

Find Investable Startups and Competitors

Search thousands of startups using natural language

Proxis (YC S24

⚠️ AI-generated overview based on web search data – may contain errors, please verify information yourself! You can claim this account with your email domain to make edits.

Executive Summary

Proxis provides an LLM distillation and serving layer optimized for low-latency, on-premise deployment of generative AI. The platform deploys AI email agents that ingest enterprise knowledge bases to auto-draft personalized, on-brand replies, only sending automatically when high confidence thresholds are met.

proxis.ai1K+
cb
Crunchbase
Founded 2024San Francisco, CA,US

Funding

$

Estimated Funding

$500K+

Major Investors

Y Combinator

Team (<5)

No team information available.

Company Description

Problem

Enterprises receive a high volume of emails that require manual drafting, personalization, and follow‑up, consuming significant time for sales, customer‑service, and operations teams. At the same time, deploying large language models (LLMs) for such tasks is expensive, introduces latency, and raises data‑privacy concerns when models are hosted off‑site.

Solution

Proxis delivers a two‑part platform that tackles both challenges. First, it provides a dedicated LLM distillation and serving layer that reduces compute costs and latency while allowing on‑premise or secure cloud deployment. Second, it builds AI email agents that ingest a company’s full email history and connected knowledge bases (CRMs, Notion, Drive, Slack, etc.) to auto‑draft replies that match the user’s tone and style. The agents only auto‑send when confidence thresholds are met, otherwise they remain as drafts for review. By combining model optimization with context‑rich automation, Proxis enables enterprises to handle inbox workloads at scale without sacrificing security or performance.

Features

LLM distillation and serving optimizations that lower inference cost and improve latency.

AI email agents that ingest full email history and company knowledge to generate on‑brand replies.

Context‑first drafting with live integrations to CRMs (Salesforce, HubSpot), Notion, Google Drive, Calendar, and Slack.

Auto‑send functionality that triggers only when the model’s confidence exceeds a configurable threshold.

Team library for shared templates and context packs, enabling consistent communication across users.

Tiered SaaS pricing (Essential $20 / seat / month) with a free 7‑day trial and higher‑tier plans offering dedicated support and SLOs.

VSCode extension for developers to call optimization functions directly.

Target Audience

Proxis targets enterprise sales, customer‑service, and operations teams that rely heavily on email, as well as developers and IT groups needing efficient, secure LLM deployment for internal applications.

Revenue Model

Proxis operates a subscription SaaS model with an Essential plan at $20 per seat per month, offering auto‑draft only, and higher‑tier plans that include auto‑send, live context connections, priority support, and dedicated account management.

Traction

Proxis launched in YC Summer 2024 and has been featured in YC Launch posts highlighting its AI email agents and LLM distillation platform. The company lists trusted customers such as Google, TCS, and Walmart on its homepage. As of November 2025, the platform offers a free trial and reports active enterprise adoption across multiple inboxes per sales rep.

Sources:
Want to add first party data to your startup here or get your entry removed? You can edit it yourself by logging in with your company domain.