Loading slides and images...
LISTEN → SEE → REASON → REMEMBER → ACT
CONCYA
An agentic AI runtime for the physical world.
We deploy place-bound AI agents into real locations. The body changes. The intelligence persists.
FOUNDER
Olaolu
2026
Pre-Seed
INTELLIGENCE WHERE TRIAGE HAPPENS
The Problem
Every confused rider, tourist, or caller hits a staff member or call center. That architecture doesn't scale. Voice is lost. Context is fragmented. Memory doesn't compound.
Hospitality
Revenue Leakage
40%
Missed
Annual Loss
$24K
per location average
Small family restaurant, two people in the kitchen—can't handle phone triage. High-intent channel with zero instrumentation.
Public Sector
Context Failure
$70K
Per Operator
Human Operator
$70K
AI Runtime
—
Daily Impact
7/10
riders left confused or misrouted (MTA-scale)
Infrastructure exists, but intelligence layer is missing.
Government & Security
Critical Latency
60s
Wasted
Missing Infrastructure
No context synthesis
No memory layer
No intelligence runtime
Manual processes under pressure. No compounding intelligence.
Concya's belief: triage should live in the environment itself. Places should listen, reason, remember, and act—so the first response happens where the problem is.
The Timing
The substrate was missing. We built it.
OpenAI, Google, and Amazon built the brain. Nobody built the body. The infrastructure exists. The runtime doesn't. That's Concya.
Sub-Second Response
<1s
Real-time voice response is now fast enough to sustain human trust. Sub-second latency makes voice feel natural, not robotic.
Economics at Scale
90%↓
Speech + reasoning can run continuously at scale. The economics finally work for infrastructure, not just demos.
Infrastructure Exists
100%
Telephony, edge compute, and APIs are programmable. The infrastructure layer exists — it just needs intelligence.
The Missing Substrate
Gap
Systems are being rebuilt for agent interaction, but lack real-world grounding. The substrate was missing. We built it.
THE INSIGHT
Before writing, before code, before APIs. Voice carried intent, context, state, and trust. We decoded it.
HIDDEN SIGNALS IN VOICE
I need to get to Newark
Where is the nearest exit?
Can I make a reservation?
What time does the train arrive?
Standing at Platform 3, rush hour
At the entrance, looking for directions
In the restaurant, checking availability
Late night, station is nearly empty
Frustrated, running late
Urgent, needs immediate help
Calm, just browsing
Confused, first time visitor
Is this system allowed to route me?
Can I trust this recommendation?
Is my data secure here?
Who has access to this information?
THE GRAIN INDEX
Extracting hidden signals from voice infrastructure
Prosody
Stress, urgency, tone
Patterns
Behavioral history
Context
Environment + location
Reliability
Trust thresholds
Outcomes
Did it work?
Each interaction compounds. Voice infrastructure becomes a self-improving intelligence asset.
Everyone sees voice as audio. We see it as the operating system for human intent.
What Makes a Concya Node
A universal intelligence loop for real-world operations. One runtime, every vertical.
Concya
Node
LISTEN
Capture voice signals
SEE
Where you are
REASON
Decide what matters
REMEMBER
Build intelligence
ACT
Execute in real world
Each loop makes the next loop smarter
One intelligence loop. Every vertical. Infinite surfaces.
THE ATOMIC UNIT OF SCALE
A Concya Node is the smallest deployable unit of real-world intelligence. It is how intelligence enters physical space, accrues memory, and produces outcomes.
One unified agentic AI that combines all five principles into a single, persistent intelligence loop.
The Five Principles (Unified)
All five principles operate as one agentic AI system. Interprets intent, applies policy, accumulates memory.
Identical across every Concya Node. Only the Body changes.
The deployment adapter that defines how humans enter the runtime. Bodies can change without retraining the runtime.
Unique node_id
Every node has a unique ID tied to a physical location. Memory belongs to places, not sessions.
Different Jobs. Same Runtime.
(Revenue Recovery)
Capture high-intent demand and convert it into bookings.
Phone line → voice ingress
(Wayfinding)
Resolve confusion at decision points in real time.
QR → visual + text ingress
(Workflow)
Expose the runtime for custom, policy-bound actions.
API → machine ingress
"All five principles unified in one agentic AI."
Concya scales by deploying nodes, not users.
This is how intelligence becomes infrastructure.
The same loop works under chaos (hospitality) and in infrastructure (public sector)
PROOF #1: HOSPITALITY
Monthly Impact Scenario: 600 calls/mo @ 3 mins/call
Before Concya
59%
calls resolved
After Concya
98%
calls resolved
What They Paid
Phone: $2
1,800 min × $0.15 = $270
420 actions × $0.50 = $210
$482
Total Bill
Value Created
Recovered revenue: $1,800
Labor saved: $360
No-shows prevented: $200
$2,360
Total Value
Net Impact
$1,878
net positive
$2,360 value − $482 cost = $1,878 saved
OPERATIONAL INTELLIGENCE
Remember Known Customers
Recognizes repeat callers, recalls order history & preferences.
Order Intelligence
Tracks phone orders, popular items, peak times, patterns.
No-Show Detection
Marks no-shows automatically, prevents future occurrences.
Result: 98% calls resolved (up from 59%). Concya handled 600 calls, recovered $2,300+ in revenue & labor savings. Net: almost $1,900 saved in one month.
PROOF #2: STATION INSTRUMENTATION
39th Ave–Dutch Kills (LIC) · 5 nodes live · QR-based, multilingual guidance
30-DAY PILOT SCOREBOARD
5
nodes
800
interactions
1,200
queries
5
languages
86%
self-serve
2.8s
response
1.5×
avg queries
TOP INTENTS (30-DAY)
Manhattan direction
22%
Transfers
18%
Next train / ETA
14%
Service delays
12%
OMNY / fare
10%
+ 4 more (accessibility, layout, safety, other)
NODE TOPOLOGY
32%
23%
21%
14%
10%
76% of questions at entrances + hub
KEY INSIGHTS CAPTURED
76%
Confusion at decision points (E1/E2/H)
Time-shaped
Weekday: direction/transfer. Weekend: exits
5 languages
Multilingual demand is operational reality
Measurable
Every interaction = node_id + intent + outcome
SPONSORED ACCESS
Rider scans → chooses "Sponsored Access" → receives time-limited token
Community sponsorship tied to mobility
Measurable, auditable pathway
New revenue stream
Need help?
SCAN
Concya Atlas
Same loop: Scan → Ask → Guide → Log → Signal
One produces revenue, the other produces signal.
Same loop. One produces revenue, the other produces signal.
"BUSINESS MODEL"
Simple pricing: per location/node/seat + action volume. Scales across surfaces (stations, hotlines, desks).
Cheapest All-in-One
$0.15/min + $0.50/intent
$482/month
600 calls/mo (Sylla pilot)
ARR Potential
$5.8K–$10.8K ARR/location
VOLUME
Phone line + minutes + actions
Usage-based. Zero upfront. Scales with call volume and intent actions (orders, reservations, inquiries).
Per node + scan volume
Custom pricing
Citywide contracts
ARR Potential
NYC: $2.4M–$4.7M ARR potential
VOLUME
Stations, kiosks, hotlines
Node-based deployment. Per-location or citywide contracts. 5–4,720+ nodes per city.
Enterprise
Post-Series A
Multi-vertical
ARR Potential
Unlimited scale
VOLUME
Multi-vertical
Platform APIs for vertical expansion and custom workflows. SDK for developers.
UNIT ECONOMICS
Usage-based pricing scales with volume. Zero upfront costs. High gross margins.
Example: Sylla pays $482/mo, gets $1,878 net value. Annualized: ~$22,500/location.
90%+
GROSS MARGIN
$482
AVG/MO
THE COMPETITIVE MOAT
The big players build generic AI. We build place-bound infrastructure.
Place-Bound AI is our defensibility. AI that belongs to a location, not a person. With location memory that compounds and policy-constrained actions for safe deployment.

The Gap
Generic Cloud Assistant
Doesn't control physical infrastructure
ChatGPT is person-bound, not place-bound. No location memory. No node-specific intelligence.
Missing
Missing: Place-bound deployment

The Gap
Data Platform, Not Runtime
Analyzes data after-the-fact, doesn't capture real-time intent
Palantir ingests existing data to inform humans. Concya operates in the moment of action (voice → decision) before data exists. Palantir is dashboard-driven; Concya is sub-second infrastructure.
Missing
Missing: Real-time voice runtime

The Gap
Device-Bound, Not Venue-Bound
Tied to hardware, not infrastructure
Alexa lives in your Echo. It doesn't belong to the restaurant, transit station, or emergency dispatch.
Missing
Missing: Venue-bound intelligence
Place-Bound Runtime
Place-Bound
AI tied to specific location (node_id), not generic cloud
Place-Memory
System learns what works at that spot over time
Policy-Constrained
Actions limited by location type (transit guides, can't transact)
IP Protected
Place-Bound AI Agent Runtime
Q1 2026 Provisional
They build AI for everyone. We build AI for places.
Place-bound memory + policy constraints + location-specific deployment = Infrastructure moat they can't easily replicate.

AMERICAN DYNAMISM
TAM: $9.83B | Target: 1-7% of ~750K US restaurants in 36 months
"We the People, We Can Do It" — Building American infrastructure, one node at a time. Each restaurant = one Concya node.
TAM
$9.83B
US Hospitality Market
TOTAL NODES
~750K
1 node = 1 restaurant
TARGET RANGE
1-7%
Penetration in 36 months
NODE PENETRATION TARGETS
1%
7,500 nodes
3%
22,500 nodes
5%
37,500 nodes
7%
52,500 nodes
LOW CASE
Per Node
$482
per month
1%
$43.4M
ARR
3%
$130.1M
ARR
5%
$216.9M
ARR
7%
$303.7M
ARR
HIGH CASE
Per Node
$900
per month
1%
$81.0M
ARR
3%
$243.0M
ARR
5%
$405.0M
ARR
7%
$567.0M
ARR
American Dynamism: Targeting 1-7% of ~750K US restaurants (7.5K-52.5K nodes) over 36 months. At $482-$900/month per node, ARR potential ranges from $43.4M-$567M. Each restaurant = one Concya node. We the people, we can do it.
THE PUBLIC SECTOR PROBLEM
Data from our MTA 39th Ave pilot reveals massive inefficiency. Infrastructure is broken where it matters most.
PILOT SITE
30-Day Pilot
KEY INSIGHT
76% of questions cluster at decision points.
Entrances, platforms, exits—exactly where infrastructure fails today.
Trips/Year (US)
9-10B
Massive scale with fragmented infrastructure. No unified intelligence layer.
Languages (NYC)
200+
NYC serves 200+ languages, but Help Points can't. Static signs fail multilingual riders.
Per Operator/Year
$70K
Human operators in glass boxes, yet 4M daily riders still face barriers. Not scalable.
Questions at Decision Points
76%
From our 39th Ave data: 76% cluster at entrances, platforms, exits—where infrastructure fails.
Public Sector / Infrastructure Gap
MTA / NYC Mayor's Office / OMNY
American Dynamism
From proof-of-concept to citywide infrastructure to Community-Supported Mobility.

Total Scale
5-20 Nodes per station
(Min: 2-5 kiosks if approved)
39th Ave Pilot (Dutch Kills). 5 QR nodes, 5,000 daily riders, 40% opt-in. Zero CapEx for city. Prove demand + capture data.
Scale to 472 NYC stations. 5-20 NFC/QR codes/kiosks per station = 2,000-4,076 Concya nodes deployed. Triggers: 70% satisfaction, 40% opt-in, $0.14/ride subsidy. Multilingual, 24/7. Self-funded.
70% satisfaction • 40% opt-in
2,000-4,076 Concya Nodes Deployed
Solve the funding gap. Brands pay for attention. Cities get infrastructure. A new revenue model.
Validated: MTA OMNY + Mayor's Office
Flipping the Deficit
$3.50
Deficit/Ride
$0.14
Subsidy/Ride
~$255K
Yield/Station/Year
Operational Gains
"Measurable ROI for the City"
The Funding Gap
Cities can't pay for new tech. Brands (SMBs, Nike, Google Maps) will pay for attention.
Aligned Incentives
Riders get free rides. Brands get placements. City gets infrastructure at zero CapEx.
BRANDS
Pay for Access
RIDERS
Get Free Service
CITIES
Get Infrastructure
"Not cheap ads. Community-supported infrastructure."
Business Model Deep Dive
Turning public transit cost centers into high-yield revenue generators through context-aware sponsorship.
"We don't just reduce costs. We create yield."
The Reality
Rider Fare
$2.90
Actual Cost
~$6.50
$3.50 Deficit / Ride
Fares cover only 45% of operating cost
Cities operate at a massive deficit.
~$5B annual shortfall (MTA)
CONTEXT
Rider at 39th Ave (Dutch Kills)
THE EXCHANGE (Choose One)
Watch 15s Ad
Opt-in Data
SPONSOR
LIC Bakery
Pays $0.50 CPCV
(Per Completed View/Lead)
TOKEN
1 Free Entry
($2.90 value)
Unit Economics
Ad Revenue
$0.50
Per View / Lead
MTA / City
70%
Concya
30%
Annual Yield
~$204M/yr at NYC Scale
~$255K per station
"Infrastructure that pays for itself."
The Origin Story
"My mum risked everything to get me and my sister to school in America, but her strong Nigerian accent cost her a job as a restaurant host. That moment made it clear: we need a voice layer that works for everyone — warm, natural, and fair."
Joint B.S./M.S. Computer Science

A.S. Comp Sci • Dean's List x4
Awards & Recognition
5 ITEMS

UNICEF Global Impact Award

Foundry Aware • Foundry Training • Application Developer • Data Scientist

Investment Opportunity
Raising $500K on SAFE to reach Seed Readiness ($3M ARR Pace)
Month 6
15% Paying Hospitality
10% Deployed In NYC Transit
The Terms
RAISE AMOUNT
$500,000
INSTRUMENT
SAFE (Post-money)
VALUATION CAP
$10M
Use of Funds
• Voice/Telephony Specialist & Infra• Goal: Sub-400ms latency & hardened engine
• "Boots on the ground" Sales (NYC)• Goal: 50 Restaurant Nodes + 5 Transit Pilots
• Legal, Compliance, Corp Setup• Goal: 18-month runway buffer
The Milestones
Progression to Series A
Month 6
$60K MRR
$720K Annualized
Action: 100 nodes, 20 locations, measurable outcome lift, retention curves, 2 paying cohorts
"Prove the 'Day 1 Painkiller' sales motion is repeatable."
Month 12
25% Coverage
NYC Metro Transit
Action: Deploy Nodes in 118 NYC Transit Hubs
"Validate the 'Sponsored Access' revenue model at scale."
Month 18
$3M ARR Pace
State Contract LOI
Action: Expand to 2nd Metro (DC/Chicago) or Enterprise Deal
"Ready to raise $4M Seed to scale the machine."
"We are raising a tight $500k to maintain discipline. We have validated the product (Sylla) and the demand (Transit). This capital allows us to hire the core technical talent to harden the runtime before stepping on the gas for the Seed round. It connects Money to Metrics: We aren't just 'hiring devs', we are buying 400ms latency."
THE VISION
Our vision: make cities intelligent at the point where triage actually happens. Long-term, global triage becomes agentic, place-aware, and eventually orbital—an intelligence layer for countries and infrastructure at scale.
Concierge nodes in every restaurant, hotel, and venue across the world.
Concierge nodes in every transit system across the world.
In every emergency dispatch in hospitals.
54 COUNTRIES • LEAPFROG SCALE
911 runs on 1960s architecture—never built for scale, multilingual, or real-time context. Most of the world has no equivalent. Across 54 African countries, no such infrastructure. Africa leapfrogs to AI-native.
FOUNDER ADVANTAGE
Pre-COVID GovTech: Nigeria Relief Tool
Active Outreach: Globacom & Tony Elumelu
When triage lives in the environment—when places listen, reason, remember, and act—that's the vision. One node, every interface. The body changes. The node stays continuous.
THE HORIZON
The coordination layer that turns local place-intelligence into shared situational awareness.
Place Nodes (Edge)
Phone · QR/NFC · Kiosks · Voice terminals
Local triage happens where the problem occurs.
Orbital Layer
Detect patterns
Coordinate escalation
Maintain continuity
Operators & Institutions
Transit Ops · City Services · Emergency Mgmt
A single operational picture across systems and borders.
Signals (not people)
Anonymized event patterns across places
Actions
Routing · Alerts · Playbooks · Resource coordination
Resilience
Works through outages, congestion, and fragmented infrastructure
Not surveillance. Not command-and-control. A shared nervous system for real-world triage.
Local intelligence becomes planetary when coordination lives above geography.