Confidential

This deck is password-protected.

Incorrect password

Skip to content
01 / 14
↑↓ Navigate or scroll
Human Intelligence
Human Intelligence

Human Intelligence

The Trust Layer for Agentic AI

Pre-Seed | $1.5M Raise | TechStars Founders
humanintel.net
The Problem

AI Agents Are Failing in Production

89% of AI agent pilots never reach production. The bottleneck isn’t capability — it’s trust.

Quality Failures
Hallucinations, wrong decisions, confidently incorrect outputs sent to customers.
Silent Degradation
Model drift, prompt rot, data shifts. Performance decays with no alerting.
No Accountability
No audit trail. No human review. No way to answer “why did the agent do that?”
Sources: Gartner 2025, Capgemini AI Agent Survey 2025
The Trust Gap

Agents Are Shipping. Nobody’s Grading Them.

40% of enterprise apps will embed AI agents by end of 2026. Quality infrastructure hasn’t kept pace.

75% of leaders say compliance & auditability are the #1 requirement
Enterprises won’t deploy agents without provable oversight. The EU AI Act, NIST RMF, and California’s SB-833 now mandate demonstrable human review.
60% of enterprises restrict agent access without human oversight
Nearly half employ human-in-the-loop controls across high-risk workflows. Demand for human QA operators is built into the adoption curve.
40%+ of agentic AI projects will be abandoned by 2027
Not because the AI doesn’t work — because there’s no governance layer. Audit trails, real-time monitoring, and quality scoring are afterthoughts.
Sources: Gartner (2025–2026), Cisco AI Security Report (2026), EU AI Act Article 14, California SB-833
Market Opportunity

$10.9B in 2026. $53B by 2030.

$2.5 trillion in global AI spending this year. Every agent deployed needs monitoring, QA, and accountability.

$10.9B
AI Agent Market
2026
$53B
Projected by 2030
46% CAGR
$492M
AI Governance Spend
2026 → $1B+ by 2030
40% of enterprise apps will embed AI agents by year-end. Governance is a $1B+ category — and it’s just starting.
Sources: MarketsandMarkets (2026), Gartner AI Spending Forecast (Jan 2026), Gartner AI Governance Report (Feb 2026)
Why Now
$575M+
raised in AI agent security in March 2026 alone. Zero dollars into quality & trust.
Promptfoo
Acquired
by OpenAI
Mar 9
Armadin
$190M
AI security · Accel, In-Q-Tel
Mar 10
Kai
$125M
Agentic AI defense
Mar 11
Bold
$40M
Endpoint AI security
Mar 12
XBOW
$120M
Autonomous offense · $1B+ val
Mar 14
Certiv
$4.2M
AI agent governance
Mar 16
RunSybil
$40M
AI pen testing · Khosla
Mar 18
Surf AI
$57M
Security ops automation
Mar 20
$575M+ poured into AI agent security. But security ≠ trust. No one is grading whether agents actually work.
Positioning

Security ≠ Trust

Security

“Is this agent allowed to do this?”

  • Access control & permissions
  • Guardrails & safety filters
  • Detects unauthorized actions

An agent can pass every security check and still hallucinate.

vs

Trust

“Is this agent doing this well?”

  • Quality, accuracy & drift detection
  • Outcome measurement
  • Catches authorized but wrong actions

Trust is what gets agents from pilot to production.

We don’t compete with security tools. We complete the stack.
The Solution

Bobby AI Agent Command Center

Human operators + proprietary platform monitoring AI agent fleets in real time.

Bobby Command Center — dashboard showing fleet status, health scores, spend tracking, live feed, and agent monitoring
01
Connect
Plug into any AI agent stack via API
02
Monitor
Real-time observation of decisions & outputs
03
QA
Automated scoring + human review of every interaction
04
Trust
Audit trails, benchmarks, compliance docs
Human operators generate labeled data that makes the platform smarter. Humans are the flywheel, not the cost center.
Business Model

Deploy in Existing Workflows. Scale Through HI’s API & Platform.

Inside Client Workflows

Human Intelligence works within existing client systems and operating processes

Adds AI review, exception handling, QA, and human-in-the-loop decisions

Low-friction deployment with recurring managed-service revenue

No rip-and-replace — we layer into what already exists

Through HI API & Platform

Clients send AI outputs and review tasks directly into Human Intelligence

HI routes work through AI + human oversight for quality and resolution

Expands into deeper integrations, higher-volume workflows, and platform revenue

Managed delivery first — deeper software integration over time

Delivery Cost
$800–1,000/mo
→
Client Revenue
$2,500+/mo per operator
→
Gross Margin
60–68%
The human layer earns trust, works inside existing workflows, and generates the quality data that strengthens the platform over time.
Market Size

Every AI Agent Needs Quality Infrastructure

TAM
$183B
Global AI Agent
Market by 2033
SAM
$28B
AI Agent QA, Monitoring
& Trust Infrastructure
SOM
$2B+
US Series A–C
Startups

SOM is deliberately narrow. We start where urgency is highest and expand into enterprise.

Bottoms-Up SOM
~12K US startups (Series A–C) deploying agents × avg 4 operators × $2,500/mo × 12mo = $1.44B addressable
Sources: MarketsandMarkets, Grand View Research, Allied Market Research, PitchBook (2025-2026)
Landscape

Everyone’s Building Agents. No One’s Grading Them.

AI Security
Armadin $190M Kai $125M Onyx $40M Bold $40M
Red-teaming, threat detection, access control, endpoint security
No operational quality, coaching, or brand adherence
BPOs
Concentrix Teleperformance TaskUs Accenture
Deploying AI agents & managing CX ops internally
No third-party AI agent QA product
Observability
Datadog Arize $70M LangSmith Helicone
Traces, latency, token usage, error rates — developer metrics
No business-outcome QA, no human oversight
CX QA Platforms
AmplifAI Oversai Level AI Enthu.AI
Scorecard QA for human agents; Oversai is AI-only, no humans
Not purpose-built for AI agent ops quality
The Trust Layer
Human QA experts + AI platform grading AI agents on brand voice, compliance, empathy & resolution — like BPO QA for the AI era
We own this quadrant.
$400M+ raised in AI agent security in one week. Zero dollars into the trust layer that grades whether agents actually work.
The Team

Operators Who Build

Melissa Powell
Melissa Powell
Co-Founder & Co-CEO
Strategy · Govt Relations · Product
  • Techstars NOLA Powered by J.P. Morgan ’24 Founder (Pocmi)
  • Woman of Impact in STEM 2025 (National Award)
  • Judge, Columbia University AI Startup Challenge
  • Draper University Hero Training (full scholarship)
  • 2x M.S. (Intl Business + Construction Mgmt) · FIU
  • Government of Jamaica Ministry of Labour & Social Security — Innovation Transformation Steering Committee
  • Delegate, Women of Color in Blockchain Congressional Hearing, Capitol Hill ’24 & ’25
TechStars Virgin MIT GBA WBAF
Matt Ford
Matt Ford
Co-Founder & Co-CEO
Technical Execution · Global Ops · Compliance
  • Founder & CEO, CommerceJet (e-commerce SaaS) & Auqli
  • Techstars Atlanta Powered by J.P. Morgan ’24 Founder (CommerceJet)
  • Scaled BPO 20→1,000+ seats, $6M+ rev yr 1 (Contexta Global)
  • Walmart International Store Planning · VP IT, Pure Global Brands
  • Built & operated teams in Manila & Cebu, Philippines, Colombia, Honduras, Nigeria, & Romania
  • PCI DSS L1, SOC 2 Type II & HIPAA compliance (enterprise BPO)
  • B.S. Management Info Systems · U of Arkansas
Greyhound SiriusXM TechStars WTIA
Joining Q2 2026
CTO / Chief AI Officer
  • Sr. Software Engineer & Data Scientist, Amazon (6yr)
  • Built AI agent creation platform
  • ML Engineering, LLMOps, Generative AI
  • Machine Learning · Stanford
Co-founders since June 2025
Melissa and Matt built Human Intelligence’s BPO operations together from day one — recruiting, training, and managing the QA workforce that powers the platform. They’ve shipped product, closed pilots, and scaled a distributed team across three continents before raising a dollar.
The Ask
$1.5M
Pre-Seed Round
Post-Money SAFE · $8M Cap
18-Month Milestones
  • $1M ARR — 3–4 enterprise clients live
  • Bobby GA release with integration marketplace
  • 30+ operators across the Philippines & Caribbean
  • Quality benchmark dataset — proprietary moat
  • Series A ready: $300K+ ARR run rate
40%
Engineering
Bobby platform, integrations, automation
40%
Operations
Human operator team, QA playbooks
20%
Go-to-Market
TechStars network, direct sales
Enterprise pilots live Bobby GA Quality benchmarks Series A
Human Intelligence

The trust layer for the AI workforce.

“You don’t trust what you can’t measure. We measure it.”

TechStars ’24 Enterprise Pilots Live $1.5M Pre-Seed 3 Continents
Seattle Headquarters
Caribbean Operations
Philippines Operations
investor@humanintel.net humanintel.net