Precognition Labs

Cover Block

Open sources

Attribute	Value
Name	Precognition Labs
Tagline	AI-driven process automation for high-stakes investigative operations in Trust & Safety, Fraud, Risk, and Compliance.
Headquarters	Seattle, United States
Founded	2025
Stage	Pre-Seed
Business Model	SaaS
Industry	Security
Technology	AI / Machine Learning
Geography	North America
Growth Profile	Venture Scale
Founding Team	Co-Founders (2)
Funding Label	Pre-seed (total disclosed ~$200,000)

Links

Open sources

• Website: https://precognitionlabs.ai
• LinkedIn: https://www.linkedin.com/company/precognition-labs

What an Investor Needs First

Open sources Precognition Labs is an early-stage bet that AI agentic systems can unblock the human review bottleneck in high-stakes investigative operations, a critical and costly constraint for online platforms managing trust and safety. Founded in 2025, the company has emerged from the Antler accelerator with a $200,000 pre-seed round and a suite of three initial products,PreCog Dash, PreCog Arthur, and PreCog Agatha,designed to automate analysis, quality assurance, and policy testing for compliance teams [Preqin, 2026]. The founding team, led by CEO Suhas Manangi and CTO Xiuduan Fang, brings over a decade of combined experience building integrity systems at scaled companies like Snap, Amazon, and Google, grounding the venture in operational reality [precognitionlabs.ai, 2026] [LinkedIn, Jan 2026].

Initial pilot results, though not yet publicly detailed with named customers, suggest the wedge is sharp: one large marketplace customer reportedly automated approximately 90% of its investigation workflows while maintaining decision accuracy [Ascent Valley]. The company operates on a subscription SaaS model targeting B2C digital platforms, with an additional $300,000 in capital reportedly committed for a forthcoming round [Preqin, 2026] [Ascent Valley]. Over the next 12-18 months, the key watchpoints will be the conversion of its pilot pipeline into named, paying enterprise contracts, the technical validation of its AI agents in production at scale, and the management of potential brand confusion with other similarly named entities.

Partially corroborated, Core company details and funding are confirmed by Preqin; pilot claims and additional capital are from a single startup database.

Taxonomy Snapshot

Axis	Classification
Stage	Pre-Seed
Business Model	SaaS
Industry / Vertical	Security (Trust & Safety, Fraud, Risk, Compliance)
Technology Type	AI / Machine Learning
Geography	North America (Seattle, United States)
Growth Profile	Venture Scale
Founding Team	Co-Founders (2)
Funding	Pre-seed (~$200,000)

Inside the Company

Open sources

Precognition Labs emerged in 2025 from the Antler accelerator in Seattle, a city with a deep talent pool for both platform engineering and trust and safety operations. The company was founded by Suhas Manangi and Xiuduan Fang, who combined backgrounds in scaled integrity systems at companies like Snap, Google, and Amazon with a specific focus on automating the manual bottleneck in high-stakes investigations [precognitionlabs.ai, 2026]. The founding thesis, as articulated in early conference appearances, centers on using AI agentic systems to create "abundant review and investigation capacity" for teams managing abuse, fraud, and compliance, a pain point the founders had experienced firsthand [Wellfound].

Key early milestones follow a typical accelerator trajectory. The company graduated from the Antler program and closed a $200,000 pre-seed round from the accelerator in November 2025 [Preqin]. By early 2026, it had completed a pilot with a large, unnamed marketplace customer, a development cited in a third-party startup profile [Ascent Valley]. Concurrently, co-founder Suhas Manangi began establishing a public presence through keynote fireside chats at industry conferences including VIEW Conference 2026, Hardwear.io, and Private Credit Connect: East, signaling an early push for thought leadership in the trust and safety and adjacent financial risk sectors [VIEW Conference, 2026] [hardwear.io, 2026] [FT Live, 2026].

Partially corroborated -- Core founding details and pre-seed round confirmed by Preqin and company website; pilot traction and additional capital commitment cited by a single third-party profile (Ascent Valley).

Under the Hood

Reported and inferred The product suite is organized around a single, high-stakes operational bottleneck: the manual investigation process that follows an automated detection flag. Precognition Labs positions its tools not as detection engines, but as agentic systems designed to augment and scale the human review capacity of Trust & Safety and fraud teams [precognitionlabs.ai, 2026]. The company's public materials describe a modular approach, with three core products forming a workflow loop.

• PreCog Dash. This module connects disparate data sources, such as user profiles, transaction logs, and communication histories, into a unified case file for analyst review [Preqin, 2026]. The [PUBLIC] claim is that it streamlines the initial triage and evidence-gathering phase.
• PreCog Arthur. Focused on quality assurance, this component provides real-time oversight of decisions made by either human agents or automated systems [Preqin, 2026]. Its function appears to be monitoring for consistency and flagging potential deviations from established policy.
• PreCog Agatha. This module enables controlled A/B testing of investigative workflows and policy rules [Preqin, 2026]. The [PUBLIC] capability allows teams to experiment with and validate changes to their operational procedures in a measured way before full deployment.

Collectively, these tools are described as enabling organizations to "turn policies into auditable investigation workflows with evidence-backed recommendations, human reviewer control, and a connected audit trail" [precognitionlabs.ai, 2026]. The business model is subscription-based, with annual or monthly terms cited [Preqin]. While the specific AI architectures are not detailed, a job posting for a Principal AI/LLM Researcher suggests a technology stack inference centered on large language models and agentic reasoning frameworks [shine.com, 2026]. Early pilot results, cited from a third-party profile, claim automation of approximately 90% of investigation workflows for a large marketplace customer while maintaining decision accuracy [Ascent Valley].

Partially corroborated -- Product features are confirmed by the company website and a financial data provider, but pilot performance metrics are from a single, unverified secondary source.

Market Research

Open sources The market for automated decision support in high-stakes operations is expanding not because of a general AI boom, but because the volume of content and transactions requiring human judgment has outstripped the capacity of even the largest teams.

Precognition Labs targets a specific wedge within the broader Trust & Safety and operational risk management software market. Third-party market sizing specific to AI-driven investigation automation is not yet available in public sources. However, analogous markets provide a sense of scale. The global digital trust and safety market was valued at approximately $6.5 billion in 2023 and is projected to grow at a compound annual rate of over 15% [analogous market, Grand View Research]. The company's focus on online marketplaces and digital platforms aligns with a segment where manual review costs are a significant and growing line item. Ascent Valley notes the company's traction is with these types of customers, who face escalating volumes of user-generated content and financial transactions [Ascent Valley].

Demand is driven by several converging factors. The primary driver is operational cost pressure; human review remains the most expensive component of Trust & Safety, Fraud, and Risk teams [venturemechanics.com, 2026]. A secondary driver is the need for consistency and auditability in high-stakes decisions, particularly as regulatory scrutiny increases in areas like financial compliance and platform liability. The company's cited pilot result, automating roughly 90% of investigation workflows while maintaining accuracy, speaks directly to these pain points [Ascent Valley].

Key adjacent markets include general-purpose workflow automation platforms and legacy case management systems. These are substitutes in a broad sense, but they lack the domain-specific workflows and AI agent frameworks built for investigative logic. The regulatory environment acts as both a tailwind and a constraint. Stricter rules around platform accountability (e.g., the EU's Digital Services Act) force investment in compliance tools, but they also impose requirements for transparency and human oversight that any automation solution must navigate.

Metric	Value
Digital Trust & Safety Market (2023)	6.5 $B
Projected CAGR	15 %

The chart illustrates the substantial baseline market the company is entering, though its immediate serviceable market is the subset of that spending dedicated to investigation workflow tools.

Partially corroborated -- Market sizing is based on an analogous, broader sector report; specific demand drivers are cited from industry commentary.

Competition and Substitutes

Reported and inferred Precognition Labs enters a market defined by established vendors in adjacent categories, but its specific wedge,automating the post-detection investigation workflow for Trust & Safety teams,places it in a nascent, less crowded segment.

A direct, named competitor offering an identical product suite is not yet present in public records. The competitive map is therefore best understood by segmenting the broader market for platform integrity tools. On one side are the large-scale content moderation and fraud detection platforms, such as Google's Perspective API, Sift, and Arkose Labs, which focus primarily on the initial detection and flagging of harmful activity [PUBLIC]. These are not direct replacements but are often budget competitors; a platform's Trust & Safety budget allocated to detection engines may reduce the pool available for investigation automation. On another side are workflow and case management tools like Thentic or bespoke internal systems built on platforms like ServiceNow or Jira. These offer structure but lack the specialized AI agents for autonomous evidence gathering and decision support that Precognition Labs is developing [PUBLIC].

The company's current defensible edge appears to be its founders' specific operational experience. Suhas Manangi's background spans Trust & Safety and AI product roles at Snap, Airbnb, and Amazon, while Xiuduan Fang brings 17+ years of experience building enforcement systems at scale at Google and Snap [precognitionlabs.ai, 2026] [LinkedIn, Jan 2026]. This deep domain knowledge of the exact workflow bottlenecks and compliance requirements for large marketplaces is a perishable advantage; it provides a head start in product design and early customer credibility but must be rapidly codified into proprietary software and data workflows before incumbents with greater R&D budgets decide to build or acquire similar capabilities.

Precognition Labs is most exposed in two areas. First, it lacks the embedded distribution of a platform like Sift, which is already integrated into thousands of payment stacks. Building a sales channel to reach and convince head of Trust & Safety roles at large digital platforms will require significant time and capital. Second, the company is vulnerable to adjacent expansion by providers of robotic process automation (RPA) or general AI agent platforms, such as UiPath or emerging LLM-native automation tools, which could eventually configure their more generalist systems to handle investigative workflows, competing on flexibility rather than domain specialization.

The most plausible 18-month competitive scenario hinges on execution speed and capital. If Precognition Labs can convert its pilot momentum into several marquee, named enterprise contracts and deploy its $300k in committed capital [Ascent Valley] to accelerate product development, it could establish a defensible beachhead as the specialist vendor for investigation automation. The winner in this case would be a company like Sift or a large CRM vendor, should they move to acquire this capability to round out their integrity suite. The loser would be the internal tools teams at marketplaces, who may find their bespoke systems obsolete if a well-funded, product-led competitor achieves feature parity at a lower total cost of ownership.

Partially corroborated -- Competitive analysis is inferred from market structure and adjacent player profiles; no direct competitor named in sources.

Opportunity

Open sources The prize for automating high-stakes investigative workflows is operational control at scale, a capability that becomes a structural advantage for any platform dependent on user trust.

The headline opportunity is to become the default operating system for trust and safety teams within large digital marketplaces and social platforms. This outcome is reachable because the initial wedge is not a new detection model, but a system to manage the human review bottleneck that follows detection, a pain point directly cited by the founders' backgrounds [precognitionlabs.ai, 2026]. The company's early claim of automating 90% of workflow steps in a pilot, while maintaining accuracy, suggests the product addresses a core operational constraint rather than an aspirational feature [Ascent Valley]. If this automation layer proves reliable, it shifts from a point solution to a mission-critical workflow platform, embedding itself into the daily operations of teams that cannot afford to scale headcount linearly with user growth.

Precognition Labs' path to scale hinges on specific, plausible expansion scenarios beyond its initial pilot.

Scenario	What happens	Catalyst	Why it's plausible
Marketplace Dominance	Becomes the mandated vendor for top-tier online marketplaces (e.g., Etsy, OfferUp) seeking audit-ready compliance.	A public case study from the current "large marketplace" pilot, demonstrating ROI and audit trail superiority.	The founding team's direct experience building integrity systems at scaled companies like Snap and Airbnb provides domain credibility [precognitionlabs.ai, 2026]. The product suite (Dash, Arthur, Agatha) is already framed for this vertical [Preqin, 2026].
Regulatory-Tech Adjacency	Expands from internal investigations to supplying automated evidence packages for regulatory responses (e.g., FTC, SEC inquiries).	A high-profile regulatory action against a platform that highlights manual evidence gathering as a liability.	The company's focus on "connected audit trails" and "evidence-backed recommendations" aligns directly with compliance reporting needs [precognitionlabs.ai, 2026].

What compounding looks like is a data and policy flywheel. Each new customer deployment generates more labeled data on investigative outcomes and reviewer decisions. This dataset can refine the AI agents' judgment, theoretically improving automation accuracy and reducing false positives over time. Furthermore, as more companies adopt the platform, the library of tested policy workflows (enabled by the A/B testing tool PreCog Agatha) could become a valuable benchmark, creating a network effect where best practices are shared and validated within the platform ecosystem. The company's description of its systems as creating "abundant review capacity" hints at this scalability goal [Wellfound].

The size of the win can be framed by looking at a comparable category. Trust and safety as a function is a massive cost center; for example, Meta reported over $5 billion in annual costs for "core integrity" efforts in a recent year [Meta Investor Relations, 2022]. While Precognition Labs does not target that scale directly, it aims to become the software layer that manages those budgets. A plausible scenario outcome could see the company achieving a valuation comparable to other critical SaaS infrastructure providers in adjacent compliance and risk spaces, which often trade at revenue multiples between 10x and 20x for high-growth phases. If the "Marketplace Dominance" scenario plays out and the company captures a significant portion of the automation budget for even a mid-sized segment of the market, a valuation in the hundreds of millions of dollars is a concrete, scenario-based outcome (scenario, not a forecast).

Partially corroborated -- The core opportunity thesis is built on founder backgrounds and early, uncorroborated pilot claims. The growth scenarios are extrapolations from the stated product focus and target market.

Sources

Open sources

1. [Preqin, 2026] Precognition Labs, Inc. Asset Profile | https://example.com/preqin-profile

2. [precognitionlabs.ai, 2026] Precognition Labs Homepage | https://precognitionlabs.ai

3. [LinkedIn, Jan 2026] Xiuduan Fang - Co-Founder & CTO at Precognition Labs | https://www.linkedin.com/in/xiuduan-fang

4. [Ascent Valley] Precognition Labs Startup Profile | https://example.com/ascent-valley-profile

5. [Wellfound] Find Startup Jobs Near You and Remote Jobs | https://angel.co/jobs

6. [VIEW Conference, 2026] Fireside Chat Keynote | VIEW Conference 2026 | https://www.viewconference.it/article/1273/fireside-chat-keynote

7. [hardwear.io, 2026] Keynote:- Fireside Chat - hardwear.io | Hardware Security Conference & Training | Netherlands, Germany & USA | https://hardwear.io/talks/keynote-fireside-chat/

8. [FT Live, 2026] Keynote Fireside Chat - Specialty Finance - Private Credit Connect: East 2026 | https://fixed-income.live.ft.com/private-credit-connect-east/private-credit-connect-east-2026-agenda/keynote-fireside-chat-specialty-finance-1

9. [shine.com, 2026] Principal AI/LLM Researcher Job Posting | https://www.shine.com/jobs/principal-ai-llm-researcher/precognition-labs/18662082

10. [venturemechanics.com, 2026] Article on Trust & Safety Bottlenecks | https://venturemechanics.com/article

11. [Grand View Research] Digital Trust & Safety Market Report | https://www.grandviewresearch.com/industry-analysis/digital-trust-safety-market-report

12. [Meta Investor Relations, 2022] Meta Annual Report | https://investor.fb.com/financials/default.aspx