1 of 20

INVESTMENT MEMORANDUM

CONFIDENTIAL

VERITUS AGENT

The autonomous AI agent for U.S. consumer lending

A voice-first, compliance-native agent platform replacing the collections call center — built by the founder who built the last one.

STAGE

Seed · Series A pending

PREPARED BY

Michael Kopitske

DATE

June 2026

VERITUS-IC / 01

2 of 20

INVESTMENT SNAPSHOT

Veritus is building the AI infrastructure layer for U.S. consumer lending

COMPANY FACTS

Founded

2024 · San Francisco, CA

Stage

Seed — Series A pending

Raised

$10.1M · February 2026

Lead investors

Crosslink Capital, Threshold Ventures

Also in

Emergence, YC, SurgePoint, Cedar, Rebel

Founders

Joshua March (CEO), Joey Stein (CTO), David Schlesinger (CSO)

Team / status

~10–15 · Live with paying customers

INVESTMENT THESIS

01

The white space is unoccupied

No funded competitor is a fully autonomous, voice-first, compliance-native agent for U.S. lending. The category is open.

02

Dual-agent architecture is durable IP

Two agents on every call — one speaks, one supervises in real time — solving the edge cases single-agent systems fail.

03

A data flywheel no rival can replicate

A licensed debt-buyer model generates proprietary dialogue data with every portfolio — software-only players can't touch it.

04

Compliance engineered in, not bolted on

FDCPA, TCPA, FCRA, GLBA at the architecture level. PCI DSS, HIPAA, SOC 2, ISO certified at Seed.

05

The team has done this before

CEO Joshua March built and sold Conversocial — the contact-center category Veritus is now disrupting.

DEAL DYNAMICS

$10.1M

Seed round · closed Feb 2026

Entry stage

Post-seed, pre-Series A

Window

Strong conviction window before enterprise traction tips the valuation

3 of 20

THE COMPANY

The AI agent that replaces the collections call center — built by the person who built the last one

Veritus Agent is building something that does not yet exist at scale: a fully autonomous AI voice agent purpose-built for U.S. consumer lending , operating across origination follow-up, collections, and hardship negotiation with regulatory compliance engineered in from day one.

The real edge is the business model underneath the technology — a licensed debt-buyer position that feeds proprietary performance data back into the AI, creating a flywheel that gets harder to replicate with every resolved account.

The core risk is execution timing . Veritus is early-enterprise in a market that demands trust from heavily regulated lenders; the pace of converting pilots into contracted ARR will determine whether the flywheel spins up before better-funded competitors pivot toward this white space. That is the central question before full conviction.

$16–18B

Combined addressable market — software + collections agency revenue

$10.1M

Seed capital raised, February 2026

70%

Recovery improvement for AWA in the first month of deployment

5 / 5

Category dimensions Veritus satisfies simultaneously — the only player that does

VERITUS AGENT · INVESTMENT MEMO

03 / 20

CONFIDENTIAL

4 of 20

THE PROBLEM

Consumer lending operations are expensive, inconsistent, and one compliance violation away from liability

01

The human agent problem

Collections and servicing remain predominantly human and call-center-driven

High turnover, rising labor costs, inconsistent script adherence

Expensive post-call compliance audits and error remediation

Cannot scale without proportional headcount — no ceiling unlocking

02

The compliance problem

CFPB 2025 FDCPA report: rising enforcement, growing complaint volume

One non-compliant call can trigger regulatory action, class action, or license revocation

Legacy platforms treat compliance as a post-call audit filter — too slow, too late

Lenders now evaluate vendors on compliance architecture, not just price or features

03

The delinquency crisis

U.S. household debt reached $18.8 trillion in Q4 2025

4.8% of outstanding debt sits in some stage of delinquency

50–80% of approved applicants never accept their offer — CAC written off

Recovery-rate improvements are immediately measurable in dollar terms

VERITUS AGENT · INVESTMENT MEMO

04 / 20

CONFIDENTIAL

5 of 20

THE SOLUTION

Veritus replaces the human agent entirely — across every channel, workflow, and conversation

A voice-first AI agent platform built for the full consumer-lending lifecycle, connecting directly to lenders' loan management systems, CRM, payment processors, and telephony via REST APIs, webhooks, and flat files.

01

Origination follow-up

Re-engages abandoned loan applications via AI voice, SMS, and email.

Turns 50–80% application abandonment into funded loans

02

Early delinquency outreach

Proactive outreach to borrowers entering early-stage delinquency.

Catches delinquency before it reaches collections

03

Collections & payment plans

Negotiates payment plans and processes payments on the live call.

Completes the transaction with no human hand-off

04

Hardship & debt recovery

Full hardship negotiation, settlement discussions, and recovery.

Highest-yield, most complex conversation category

Channels

Voice · SMS · Email · Chat, unified

Live

Fintechs · major loan servicer · UK bank

Certified

PCI DSS · HIPAA · SOC 2 · ISO

VERITUS AGENT · INVESTMENT MEMO

05 / 20

CONFIDENTIAL

6 of 20

TECHNOLOGY · DUAL-AGENT ARCHITECTURE

No competitor has built what Veritus built — two AI agents running simultaneously on every call

Primary Agent

Conducts the live borrower conversation.

Handles voice delivery, natural language, and negotiation

Adapts dynamically based on borrower responses

Completes payments, plan agreements, and escalations on the call

REAL-TIME INTELLIGENCE

←

guidance fed back mid-call

Supervisory Agent

Listens in real time to the entire conversation.

Evaluates compliance posture mid-call

Detects edge cases: bankruptcy, disputes, attorney representation, cease-and-desist, deceased-borrower signals

Feeds recommendations and risk assessments back to the primary agent

"We use one AI agent speaking to the customer and a separate AI agent behind the scenes, listening in on the conversation and making assessments."

— Joshua March, CEO · FinTech Futures, February 2026

WHY THIS IS A MOAT

Re-architecting the inference pipeline, managing latency between two live models, and training both on lending-specific compliance. A competitor starting from scratch cannot replicate this in under 12–18 months.

VERITUS AGENT · INVESTMENT MEMO

06 / 20

CONFIDENTIAL

7 of 20

THE DATA FLYWHEEL

Not just a software vendor — a licensed debt buyer building an advantage no software-only rival can access

↻

DATA FLYWHEEL

01

Acquire debt portfolios

02

AI agents operate them

03

Proprietary dialogue data

04

Higher recovery → more portfolios

REVENUE STREAM A · SOFTWARE

Platform licensed to lenders, servicers, and collections agencies. SaaS subscription + usage (per call/message) + performance (% recovery uplift).

REVENUE STREAM B · DEBT BUYING

Veritus acquires and operates its own debt portfolios on the platform. Direct recovery revenue plus proprietary training data no competitor can buy.

A software-only competitor can build a dual-agent architecture and a compliance engine. No software-only competitor can generate the proprietary dialogue data that comes from operating real debt portfolios. The gap widens every quarter — not narrower.

VERITUS AGENT · INVESTMENT MEMO

07 / 20

CONFIDENTIAL

8 of 20

MARKET OPPORTUNITY

Entering a $16–18B addressable market at the exact moment AI is replacing legacy software

The AI layer in debt collection is growing at 16.9% CAGR — nearly 2× the base software market rate. Value is accruing to AI-native platforms, not legacy workflow tools.

MARKET SEGMENT

2024 SIZE

PROJECTED SIZE

CAGR

VERITUS RELEVANCE

Debt Collection Software

$5.34B

$9.27B by 2030

9.6%

Core wedge — collections & recovery

Loan Servicing Software

$3.28B

$7.11B by 2032

10.17%

Adjacent — post-origination servicing

AI for Debt Collection

$3.34B

$15.9B by 2034

16.9%

AI-native layer Veritus directly competes in

Digital Lending Platforms

$10.55B

$44.49B by 2030

27.7%

Broadest TAM, including origination

U.S. Collections Agency Revenue

$13.6B / yr

—

Second revenue pool via debt-buying

$16–18B

Combined addressable market

Grand View Research · Fortune Business Insights · IBISWorld, 2023–25

ADJACENT EXPANSION — SAME PLAYBOOK

Healthcare RCM

$343.78B · 11.12%

HIPAA already certified

Insurance Servicing

$736.8M · 24.4%

Compliance-heavy collection

Automotive Finance

$295.13B · 7.4%

Same workflow structure

VERITUS AGENT · INVESTMENT MEMO

08 / 20

CONFIDENTIAL

9 of 20

WHY NOW · SIX TAILWINDS

Six concurrent forces are creating immediate, auditable demand for exactly what Veritus built

01

Delinquency crisis

U.S. household debt at $18.8T , 4.8% in delinquency. Every recovery improvement is measurable — making ROI auditable to any CFO.

02

Labor cost pressure

Rising agent costs, high turnover, inconsistent compliance. Lenders need an infrastructure replacement, not a supplement.

03

Funnel abandonment

50–80% of approved applicants never accept. AI outbound re-engages them — turning existing CAC into funded loans.

04

Regulatory intensity

CFPB enforcement and complaints up. Lenders now require compliance architecture, not promises. Veritus passes legal review from day one.

05

Borrower preferences

Borrowers increasingly prefer digital, self-service contact. Veritus unifies voice, SMS, email, and chat in one system.

06

Legacy fragmentation

FICO, Experian, TransUnion, Fiserv, nCino, Finastra — none voice-first AI-native. One player can displace several legacy contracts at once.

VERITUS AGENT · INVESTMENT MEMO

09 / 20

CONFIDENTIAL

10 of 20

COMPETITIVE LANDSCAPE

No funded competitor occupies Veritus' position — the category is unoccupied

COMPANY

VOICE-FIRST

LENDING-NATIVE

COMPLIANCE-NATIVE

FULLY AUTONOMOUS

OWNS PORTFOLIOS

Veritus Agent

✓

Observe.AI · $214M

✓

Partial

✗

copilot

✗

TrueAccord · ~$77M

✗

✓

Partial

✗

InDebted · ~$20M

✗

✓

Partial

✗

Yellow.ai · $102M

✓

✗

Partial

✗

Proximitty · YC

?

Partial

?

✗

Legacy IVR / Software

✓

Partial

✗

✓ Yes · Partial · ✗ No · ? Unknown — raised figures are reported totals

THE CRITICAL DISTINCTION · OBSERVE.AI

Observe.AI ($214M, SoftBank Vision Fund) is the most-funded adjacent player — and a fundamentally different product. It is a copilot that coaches human agents; the human agent is still required. Veritus replaces the agent entirely — a lender using it does not pay for the call center.

THREAT SUMMARY

HIGH

Observe.AI, if it pivots to autonomous — requires architectural rebuild

MED

TrueAccord, InDebted, Proximitty — digital-first but not voice, or early-stage

LOW

Horizontal platforms — not compliance- or lending-native

VERITUS AGENT · INVESTMENT MEMO

10 / 20

CONFIDENTIAL

11 of 20

TRACTION & PROOF POINTS

Live in production with enterprise customers — the technology risk has already been retired

LIVE DEPLOYMENTS

Platform went live Late 2025

Fintechs — fastest procurement; the commercial wedge

Major U.S. loan servicer — enterprise validation

UK bank — international optionality signal

AWA — named, published case study

Stork Club — confirmed customer

THE AWA CASE STUDY

70%

improvement in debt recoveries within the first month of deployment.

AI agents responsible for roughly half of every dollar collected.

Published case study · AWA, Early 2026 — live production, not a pilot

ENTERPRISE CERTIFICATIONS

✓

PCI DSS

✓

HIPAA

✓

SOC Type II

✓

ISO

All four achieved at Seed. Competitors starting today are 12–18 months from the same conversation.

For investors The technology risk priced into Seed-stage AI is substantially lower here. Dual-agent architecture is live, compliance is in production, enterprise customers are processing real borrower conversations, and the data flywheel is already running.

VERITUS AGENT · INVESTMENT MEMO

11 / 20

CONFIDENTIAL

12 of 20

THE TEAM

The only team built from inside the industry Veritus is disrupting

At Seed, founder-market fit is the single most important team signal — Veritus has it in an unusually compelling form.

Joshua March

CO-FOUNDER & CEO

Conversocial → Verint

SCiFi Foods

Built contact-center software for a decade, watched the incumbent acquire it, then built the AI-native replacement. Direct founder-market fit — not pattern-matched.

Joey Stein

CO-FOUNDER & CTO

Divvy Homes

Fintech infrastructure engineering: regulated financial data, payment processing at scale, compliance-adjacent data handling from day one.

David Schlesinger

CO-FOUNDER & CSO

Divvy Homes

Bridges the software platform and the licensed debt-buying model. A CTO and CSO from day one treats technical leadership and strategy as distinct functions.

Tim Humphrey

Head of Collections

Best Egg · 10 yrs

Ran collections ops at a major consumer lender for a decade.

Stanley Lau

Head of Risk & Pricing

Goldman Marcus

LendingClub

Portfolio risk and pricing for the debt-buying model.

Ronnie Momen

Strategic Adviser

LendingClub · CCBO

C-suite relationships inside the target customer base. Shortens sales cycles.

Caleb Sima

Security Adviser

Robinhood CSO

Databricks CSO

Bank-grade security at scale — adds weight to SOC 2 and PCI.

VERITUS AGENT · INVESTMENT MEMO

12 / 20

CONFIDENTIAL

13 of 20

INVESTMENT THESIS

Five reasons Veritus is the vertical AI fintech investment the thesis was built to find

01

The category is unoccupied and the window is narrow. No funded competitor satisfies all five dimensions. Veritus is already in production — every quarter the head start and proprietary-data gap widen.

02

The data flywheel is the deepest moat in vertical AI. The debt-buyer model builds proprietary training data from the first portfolio, independent of software revenue — and no acquirer can buy this asset.

03

Compliance architecture is a buyer requirement only Veritus meets. PCI, HIPAA, SOC 2, and ISO at Seed clear a bank's vendor-risk review today. A competitor beginning now is 12–18 months behind.

04

Founder-market fit is as strong as this category offers. March built, scaled, and sold the legacy contact-center category Veritus now replaces. That institutional knowledge took years and cannot be accelerated.

05

A direct match to the vertical AI thesis. Application-layer vertical AI, proprietary data moat, first-mover, deep LMS/CRM/telephony integrations, fintech, Seed/Series A entry, U.S.-headquartered.

"The biggest AI winners will be companies that build tailor-made AI stacks for specific industries."

Veritus: built exclusively for consumer lending. Not adapting a horizontal tool. Every product and compliance decision is specific to lending workflows.

VERITUS AGENT · INVESTMENT MEMO

13 / 20

CONFIDENTIAL

14 of 20

RISK & MITIGANTS

The risks are real, mostly knowable, and in most cases already mitigated

15 of 20

AREAS FOR FURTHER DILIGENCE

Five Priority 1 conversations before full conviction

QUESTION

WHY IT MATTERS

SOURCE

Who owns enterprise sales, and what does the GTM org look like?

Determines whether the sales motion can scale to Series A without a critical hire gap

Founder call

What are current ARR, MRR, ACV, and gross margin?

Core financial proof of product-market fit; not publicly disclosed

Data room / founder

Build vs. buy on the underlying voice AI models?

Governs defensibility of the core tech as foundation models improve

Technical co-founder

Collections licensing — which states, on what timeline?

Directly governs when the data flywheel can scale nationwide

Founder / legal

Kasia Gora's background and operational scope?

COO is a critical function given dual software + debt-buying operations

Founder call

If these five conversations confirm what the public evidence suggests, the investment case is compelling. The diligence is confirmatory, not exploratory.

VERITUS AGENT · INVESTMENT MEMO

15 / 20

CONFIDENTIAL

16 of 20

RETURN SCENARIO ANALYSIS

The entry point is attractive and the exit paths are multiple

Financial metrics are not publicly disclosed — analysis based on VC-standard stage inference, comparable company analysis, and documented market dynamics.

BUSINESS MODEL

Software platform

SaaS + usage — subscription, per-call/message, % recovery uplift

Licensed debt buying

Principal revenue — portfolio acquisition price vs. recovery rate

Combined

Hybrid — software margin + recovery margin

COMPARABLE COMPANY CONTEXT

Conversocial

Seed → Exit

Acquired by Verint

TrueAccord

Series B/C

~$77M raised; strategic value to lenders

Observe.AI

Growth

$214M raised; strategic M&A candidate

FICO / Experian

Public

Incumbents being disrupted

Path 01

Strategic acquisition

Contact-center incumbents (Verint, NICE, Genesys) or fintech infra (FIS, Fiserv, Jack Henry) acquire Veritus as an AI-native lending layer. March knows this buyer set from the Conversocial exit.

Path 02

Financial sponsor / growth

At Series B/C with proven ARR and flywheel metrics, Veritus enters growth-equity range at a premium to pure-software multiples, given the hybrid model and data asset.

Path 03

Public markets / standalone

If the category tips and Veritus leads, the standalone business has comparables in both fintech infrastructure and AI software — premium multiples in each.

Estimated runway: 12–24 months from the February 2026 close, based on $300K–$800K/month burn inference for a 10–15 person team with AI infrastructure costs.

VERITUS AGENT · INVESTMENT MEMO

16 / 20

CONFIDENTIAL

17 of 20

CONCLUSION

Veritus is the vertical AI fintech investment the thesis was built to find

01

The opportunity

Fully autonomous AI agents for U.S. consumer lending — a $16–18B addressable market growing 9–17% CAGR, at the inflection point where voice AI is production-ready for regulated financial conversations. The category is not yet won.

02

The position

The only company satisfying all five dimensions of the category simultaneously. Live in production. Enterprise certified. Data flywheel already running. A founder who has done this before. The window is open — but will not stay open indefinitely.

03

The ask

Resolve the five P1 diligence items in direct conversation with the founding team. If those conversations confirm what the public evidence suggests, pursue active participation in the Series A syndicate.

VERITUS AGENT · INVESTMENT MEMO

17 / 20

CONFIDENTIAL

18 of 20

APPENDIX A

Deep workflow integration creates durable switching costs

LMS

Loan management systems

CRM

Customer platforms

Payments

Payment processors

Telephony

Call infrastructure

REST APIs

Webhooks

Flat files

Real-time exchange

⇅

Veritus Agent Platform

Replacing Veritus requires replacing the entire borrower communication infrastructure. That is not a switching cost — it is a switching barrier.

VERITUS AGENT · INVESTMENT MEMO

18 / 20

CONFIDENTIAL

19 of 20

APPENDIX B

Bank-grade security at the Seed stage

CERTIFICATION

WHAT IT COVERS

WHY IT MATTERS

PCI DSS

Payment card data security

Required to process payments on live calls

HIPAA

Health information privacy

Enables healthcare revenue-cycle expansion

SOC Type II

Internal security controls audit

Required by enterprise banks and servicers

ISO

Information security management

International standard for enterprise vendor approval

All four achieved at the Seed stage. Competitors starting this process today would not complete it for 12–18 months.

VERITUS AGENT · INVESTMENT MEMO

19 / 20

CONFIDENTIAL

20 of 20

APPENDIX C

The quantitative case for why now

INDICATOR

DATA POINT

SOURCE

U.S. household debt (Q4 2025)

$18.8 trillion

NY Federal Reserve

Debt in delinquency

4.8%

NY Federal Reserve

AI for debt collection CAGR

16.9% through 2034

Grand View Research

Loan application abandonment

50–80% of approved

Industry benchmark

Consumers paying a premium for functional benefits

86%

Industry research

CFPB enforcement trend

Increasing

CFPB 2025 Annual Report

VERITUS AGENT · INVESTMENT MEMO

20 / 20

CONFIDENTIAL