Table of Contents

Aside Investment Report

Aside Investment Report

Category: AI browser / agentic web automation (consumer and team SaaS)

Company Stage: YC-backed seed stage (Fall 2025 batch), post-viral-launch, ~5–8 employees

Founder or Founders: Jun Kim (CEO), Chanhee Lee (CPO), Sanghun Lee (CTO)

Headquarters: San Francisco, California (founding team previously in Seoul)

Funding: Y Combinator plus angel investors including Dane Knecht (Cloudflare); round size and valuation not publicly disclosed

Business Model: Freemium consumer SaaS with credit-metered usage tiers and an enterprise plan

Product Hunt Launch Date: Most recent launch the week of September 17, 2026 (initial browser launch June 23, 2026)

Report Date: September 17, 2026

Investment Metric	Assessment
Venture Potential	72/100
Unicorn Path	Plausible
Valuation Attractiveness	Not Assessable
Evidence Confidence	60/100
Final Decision	DD

Executive Summary

Aside is a Chromium-based browser rebuilt for AI agents. Instead of relying on API integrations, its agent uses websites directly — signing into accounts, sending messages and payments, operating internal tools, and working with local files — with human confirmation required for sensitive actions like sending or paying. It differentiates on a password-manager architecture (“Aside Vault”) that autofills credentials without exposing them to the agent, local-by-default memory built from browsing history, and a bring-your-own-subscription model that lets users pay with existing ChatGPT or Claude plans. aside

The initial customer is the individual power user — developers, founders, and knowledge workers who live in a browser and want recurring web work automated. The strongest positive signal is verified viral adoption: after launching on June 23, 2026, the launch post drew 1.6–3 million views and the product crossed 20,000 users within two weeks, generating enough revenue that the team shut down their previous product (an AI sales copilot). The company also claims #1 performance on three browser-agent benchmarks against OpenAI, Anthropic, and Browser Use. aside

The team is unusually credible for this stage: all three founders were early employees at Airbridge, which they describe scaling from $0 to $30M ARR, and the company persisted through five pivots and seven YC applications before acceptance. eomag

The most important concern is that AI browsers are becoming a graveyard of well-funded attempts, and the largest platforms — OpenAI (Atlas), Perplexity (Comet), The Browser Company (Dia), and Anthropic’s own agent offerings — are competing directly with vastly more capital. Retention, the metric that killed most AI browsers, is not disclosed. aside

Final decision: DD. The market is large enough for venture-scale outcomes, the team and early adoption clear a diligence bar, but revenue, retention, and unit economics require verification before any commitment.

Product Overview

Customer problem: Existing AI agents and AI browsers stop at login screens, cannot complete multi-step work on real authenticated sites, and interrupt the user for context repeatedly; most “fail when the work gets real”. aside

How it works: Aside is a full Chromium browser with an agent harness built in. Users assign tasks; the agent navigates logged-in sites directly, with credentials autofilled by the Vault password manager and never exposed to the model. Sensitive actions (payments, posts, messages) wait for explicit human confirmation. Memory derived from browsing history stays local. aside

Core features: agent tabs and routines (scheduled tasks); Aside Vault password manager with hardware-backed encryption, scoped access, and audit logs; local memory; Ultrabrowse deep research (Pro tier); remote control via channels; cloud handoff. aside

Target users: individual power users and technical professionals; enterprise tier targets teams running agents on shared company accounts. aside

Pricing: Free ($0, 500 credits/month, 3 routines); Pro $20/month (3x usage, unlimited routines, channels); Max $200/month (40x usage); custom Enterprise (seats, shared vaults, API-initiated agent runs). Users can bring existing ChatGPT/Claude subscriptions or API keys. aside

Primary benefit: complex, recurring web work — outreach, payments, internal-tool operations, document work — completed rather than suggested.

Replaces: AI browsers that only assist (Dia, Comet, Atlas) and integration-based agents (Zapier-style automations). aside

Founder and Team Assessment

Verified through YC’s company page and independent profiles: Jun Kim (CEO) was a founding member of Airbridge from age 17 and founded a developer community now exceeding 200,000 members; Chanhee Lee (CPO) was a founding engineer at Airbridge and started a Microsoft-backed nonprofit at 15; Sanghun Lee (CTO) built an app with over 1 million downloads. The Airbridge $0-to-$30M-ARR claim is founder-reported but consistent across YC profiles. Forbes profiles Jun Kim as CEO and describe the early-technical background. ycombinator

Team size shows a minor conflict: YC’s jobs page lists 5 employees (July 2026) while the YC company page lists 8 — we treat it as 5–8 with active hiring, including a founding designer role posted in July 2026 at $150K–$230K with 0.5–1.5% equity. The eomag interview independently corroborates the pivot history: seven YC applications, five pivots from the original sales-copilot product, and the browser shipped in five months by five people. Founder-market fit is strong — the team is deeply technical, product-craft-obsessed, and demonstrated unusual persistence. Consumer-distribution experience is less proven; Airbridge was B2B adtech. ycombinator

Founder Assessment: Proven technical operators with genuine scale-up experience and exceptional persistence; consumer execution remains the open question.

Market Opportunity

The initial segment is narrow: technical power users who would switch browsers to automate authenticated web work — a population in the low millions globally (developers, growth operators, founders, analysts). At $20/month Pro pricing, willingness to pay among this group is credible, and the $200/month Max tier and enterprise plan give a path to $2,000–$20,000+ ACV for teams running agents on shared accounts. aside

Bottom-up: 50,000 paying Pro users ≈ $12M ARR; 400,000 ≈ $96M ARR — achievable within the global power-user population if retention holds. The category itself has demonstrated venture-scale outcomes and valuations: Perplexity’s Comet, OpenAI’s Atlas, and The Browser Company’s Dia are all funded at unicorn-plus levels. The platform-shift thesis — the browser becomes the operating layer for AI agents — is the actual prize, and it justifies venture interest despite crowded competition. Timing is favorable now because agentic web use has crossed from novelty to daily workflow in 2026. aside

Traction and Growth Signals

Viral launch (June 23, 2026): launch post reached 1.6 million views per the founding team’s interview, and the company’s own job posting cites 3M+ views and 20,000+ users within two weeks — both company-reported, with the figures differing; we treat 20,000+ users as the more conservative anchor. ycombinator
Revenue: company states it was “enough” to shut down the previous product; not quantified or independently verified. eomag
Benchmarks: #1 on Online-Mind2Web, BU-Bench-V1, and Odysseys, with a claimed 99.0% score versus Browser Use 97.7%, GPT-5.4 92.8%, Claude Opus 4.8 84.0%, and ChatGPT Atlas 70.0% — company-reported. aside
Product Hunt (September 2026): 5.0 rating, 1 review, ~249 followers; comments from multi-week daily users comparing it favorably to Dia, Comet, and Atlas. producthunt
Hiring: active founding-designer search and community engagement. ycombinator

Missing: revenue, free-to-paid conversion, retention, active users after the viral spike, and CAC. A viral June launch with a small September relaunch is consistent with either sustained momentum or a need to re-ignite attention — the data does not distinguish.

Traction Assessment: Credible early adoption and usage depth signals, but retention and revenue are unverified.

Competitive Position

Competitor	Type	Notes
OpenAI ChatGPT Atlas	Direct, platform	Claimed 70.0% on the benchmarks Aside cites aside
Perplexity Comet	Direct	Massive installed base and capital
The Browser Company (Dia)	Direct	Arc successor; incumbent AI-browser brand; users cite profile-swiping gaps in Aside producthunt
Anthropic agent / Browser Use	Direct	Compared directly in launch claims aside
Opera Neon, Fellou, Radius	Direct	Crowded field of AI browsers extruct
Zapier / integration agents	Indirect	Require APIs; Aside uses sites directly aside

Differentiation: using logged-in sites directly rather than integrations, the Vault credential architecture, local memory, and claimed benchmark leadership are real, current advantages. The honest answer to “what if OpenAI ships this natively within six months” is that OpenAI already ships a browser — Aside’s counter is speed, craft, privacy architecture, and being model-agnostic (BYO subscription). That is a real but narrow moat; switching costs for browsers are historically low. aside

Defensibility Assessment: Low-to-Medium — strong current product differentiation, weak structural moat.

Business Model and Economics

Revenue is credit-metered SaaS across $0/$20/$200 consumer tiers plus enterprise contracts. Key economics: aside

BYO-subscription design shifts inference costs to users’ existing ChatGPT/Claude plans — a materially better margin structure than inference-burning agent startups. aside
Free tier (500 credits/month) creates a conversion funnel whose rate is unknown.
Vault security architecture (Secure Enclave, post-quantum encryption claims) and audit logs are enterprise-grade features supporting the team plan. aside
Unknowns: credit-to-cost ratio, gross margin, whether $20/month covers heavy agent usage, churn among viral adopters, and enterprise pipeline.

If usage growth is metered against credits priced above infrastructure cost, revenue can scale faster than costs — but this must be verified.

Unicorn Path

Assume a 12x ARR multiple, appropriate for a high-growth consumer AI product with network effects and category leadership claims. Required revenue: $1B ÷ 12 ≈ $83M ARR. At blended pricing (mostly $20 Pro with some $200 Max and enterprise), this implies roughly 300,000–400,000 paying users, or materially fewer if the enterprise tier (teams running agents on shared accounts, $2,000–$20,000+ ACV) becomes the revenue core — e.g., 5,000–10,000 team accounts. aside

Comparable category outcomes (Perplexity, The Browser Company) demonstrate that browser-layer AI companies can reach multi-hundred-million and multi-billion valuations. Required strategic steps: prove retention post-viral spike, build repeatable distribution beyond launch moments, scale the enterprise tier, and likely raise a large round to compete against platform incumbents. The path exists and is being walked by peers; whether Aside wins it is the core bet.

Unicorn Path: Plausible.

Valuation Assessment

Funding is disclosed only as “Y Combinator and angels such as Dane Knecht”. No round size, SAFE cap, post-money valuation, or current fundraising status is public. With no verified revenue or financing terms, no responsible valuation range can be constructed. ycombinator

Valuation Attractiveness: Not Assessable. Required information: current ARR and growth, retention, gross margin, round size and terms, post-money valuation, investor ownership, and any liquidation preferences.

Key Risks

Platform competition — OpenAI, Perplexity, and The Browser Company have vastly greater capital and distribution in the same category. aside
Retention risk — the historical failure mode of AI browsers; no cohort data disclosed.
Benchmark claims are self-reported — 99.0% on third-party benchmarks is company-published, not independently audited. aside
Unverified revenue — “enough to shut down the previous product” is unquantified. eomag
Security and liability exposure — an agent that makes payments and sends messages with stored credentials carries concentrated operational and trust risk if a mistake occurs.
Small team scale — 5–8 people against platform incumbents; key-person risk across three founders. ycombinator
Low browser switching costs — habitual lock-in is weak in this category.
Viral-dependent distribution — two launch moments in three months suggests acquisition is episodic rather than repeatable.

Final Assessment

Venture Potential: 72/100

Category	Score
Market Size and Expansion Potential	17/20
Traction and Growth Evidence	12/20
Founder and Team	12/15
Product Strength	8/10
Distribution Potential	12/15
Business Model and Economics	6/10
Defensibility	5/10
Total	72/100

Strongest elements: a genuine platform-shift market with unicorn precedent, a proven founding team, credible viral adoption, and a differentiated privacy/credential architecture with a favorable BYO-model cost structure. Weakest elements: low structural defensibility against platform incumbents and undisclosed retention/economics.

Evidence Confidence: 60/100

Verified: founder identities and backgrounds (YC page, Forbes profiles), YC F25 backing, team size range, product existence, pricing, and job listings. Company-reported: 20,000 users, view counts, revenue sufficiency, and benchmark scores. Unknown: revenue, retention, conversion, CAC, gross margin, round terms, and valuation. The view-count discrepancy (1.6M vs 3M+) and team-size conflict (5 vs 8) are minor but noted.

Final Decision: DD

Venture Potential of 72/100 sits in the venture-backable band, the Unicorn Path is Plausible with category precedent, and the team and early traction are credible enough to justify formal diligence. Valuation is Not Assessable, which rules out Invest but not DD. Diligence should focus on retention after the viral spike, revenue and credit economics, security architecture, and the defensibility question against platform browsers.

Upgrade Conditions

Verified retention: >40% of June cohort still weekly-active at six months
Disclosed ARR trajectory toward $1M+ with conversion data
Enterprise tier signings (teams on shared vaults)
A priced round with credible investors at sane terms
Independent (non-company-run) benchmark validation

Downgrade Conditions

Retention collapse after the launch spike
OpenAI or Perplexity shipping authenticated-site agent automation natively
Security incident involving the Vault or payment actions
Benchmark claims contradicted by independent testing
Founder departure or stalled product iteration

Questions for Further Diligence

What is current MRR, and what share comes from Pro, Max, and enterprise?
What were week-4, week-12, and week-24 retention rates of the June launch cohort?
What is free-to-paid conversion at 500 credits/month, and what does the average paying user consume?
What is gross margin per Pro user under BYO-subscription versus Aside-managed inference?
How are the benchmark results produced — fully independent harnesses or company-run evaluations?
What is the security model for payment execution and Vault access, and has it been third-party audited?
What did the angel round total, at what valuation, and is a priced round planned?
How is the team splitting between consumer browser and enterprise agent platform going forward?
What is the CAC of users acquired outside the two launch moments?
What is the roadmap for Windows, and does the macOS/Secure Enclave architecture gate the market?
How does Aside respond if OpenAI restricts third-party browsers’ access to ChatGPT sessions?
What is the plan if Browser Use or an open-source harness matches the benchmark performance at zero cost?