Aedilic

What do they actually do

Aedilic builds nonescape, an open‑source detector for AI‑generated images and deepfakes. It’s designed to run in the browser or Node, with inference code and models that teams can run locally or on‑prem. The library is available on GitHub and npm, with install and usage examples aimed at real‑world pipelines (GitHub, npm, site).

Today, the product is a drop‑in detection component rather than a managed service. It targets integration into upload flows, KYC checks, moderation dashboards, and security tooling where transparency, auditability, and local execution are important (GitHub, YC).

Who are their target customer(s)

• Platform/content engineers at social networks, marketplaces, and image hosts: Need a detector they can drop into upload or client‑side flows as AI‑generated images scale and erode user trust; want a browser/Node library with straightforward integration and on‑prem options (GitHub, npm).
• Fraud and KYC teams at banks and fintechs: Concerned that synthetic faces/videos bypass identity checks; lab‑tested detectors often fail in live flows, so they need local, auditable tools that plug into verification stacks (DarkReading, GitHub).
• Moderation and newsroom fact‑check teams: Must quickly assess image authenticity for takedowns and reporting; current detectors can be inconsistent on real‑world content, so they prefer an open‑source tool they can audit and run on‑prem (Gracker, GitHub).
• Corporate security / SOC teams: Face targeted social‑engineering with synthetic media; need detection that integrates into alerts and forensics and allows inspection of false positives/negatives—open‑source is favored over black‑box vendors (YC, GitHub).
• Researchers, NGOs, and government forensics labs: Require transparent, reproducible methods that run locally due to privacy and chain‑of‑custody needs; want models and inference code they can inspect and adapt (GitHub, npm).

How would they acquire their first 10, 50, and 100 customers

• First 10: Direct outreach to target engineering, moderation, and fraud/KYC teams via GitHub, developer communities, and email; offer a time‑boxed, in‑production pilot with hands‑on integration help to drop the open‑source detector into an upload or verification flow (GitHub, npm).
• First 50: Publish case studies and reproducible examples from early pilots, create how‑to guides for common stacks, and work with adjacent vendors (ID providers, moderation tools) for referrals/bundles; focus messaging on reductions in field failures that matter to KYC/fraud teams (DarkReading).
• First 100: Ship self‑serve installs plus paid support/managed on‑prem and short SLA contracts; add a small DevRel function to triage issues and convert high‑touch pilots into recurring deals and partnerships, leveraging open‑source provenance and YC credibility in procurement (GitHub, YC).

What is the rough total addressable market

Top-down context:

Aedilic sells into the detection slice of content moderation and KYC/identity budgets. Detection‑focused reports indicate hundreds of millions today, scaling to low‑single‑digit billions by 2030–2034, with adjacent markets like content moderation and KYC in the tens of billions (MarketsandMarkets, Mordor Intelligence, Grand View Research, 360 Research Reports, Yahoo/MarketsandMarkets).

Bottom-up calculation:

If ~1,500 likely buyers (platforms/marketplaces, fintechs/banks with KYC, and media/security orgs) can justify a dedicated detection component at $50k–$150k/year for support/on‑prem licensing, the bottom‑up core TAM is roughly $75M–$225M annually, with upside as adoption and per‑account spend grow.

Assumptions:

• 500 platforms/marketplaces with UGC images, 800 fintechs/banks with KYC, 200 media/security orgs as initial reachable buyers.
• Average annual contract for support/on‑prem license of $50k–$150k.
• Focus is on image/deepfake detection component, not full content moderation or KYC spend.

Who are some of their notable competitors

• Sensity AI: Commercial synthetic‑media detection and monitoring platform (images, video, audio) with APIs and on‑prem; suited to enterprises that want a turnkey service rather than an open‑source library.
• Truepic: Tamper‑evident capture and verification (SDKs/services) focused on preventing manipulated media at the point of capture, often for insurers, lenders, and marketplaces.
• Reality Defender: Enterprise real‑time deepfake detection (APIs/tools) for meetings, media, and platforms; emphasizes managed detection and fast integration.
• Serelay: Trusted capture and authenticity checks that add verifiable metadata and provide verification services—aimed at provenance and chain‑of‑custody use cases.
• Deepware: Deepfake scanning and forensic tools with public and enterprise/on‑prem options; positioned as a ready‑to‑use service rather than a drop‑in open‑source library.