Investors · Series Seed · open until close

Every drone starts with a conversation.
We are building the room where it happens.

Drone hardware grew 18× in a decade. The software that designs it did not. DRONA is the AI-native design environment for every flying machine — the layer every builder, operator, and fleet will live inside by 2030. We are raising our seed round. This page is the room.

One-liner · We shorten the distance between a problem and a machine that solves it — from 22 hours to 9 minutes.
The vision

Drone design is the next CAD. We are the Figma of it.

Every generation of hardware gets a new design layer. Mechanical got AutoCAD. Electrical got Altium. Chips got Cadence. Interfaces got Figma. Real-time got Unity. Each layer collapsed a multi-week artisanal process into a multi-hour conversation — and each layer grew a $10B+ category beneath it.

Drones do not have that layer. They have fifteen browser tabs, a forum thread from 2019, and a pile of datasheets. DRONA is the layer. Describe the job — watch the machine assemble — ship a validated build. One conversation. One artifact. One standard.

The drone is only the first machine. The architecture — chat as interface, physics as authority, catalog as truth — extends to every buildable thing in the sky, on the water, underground, and eventually on other planets. This is the conversation layer for physical systems. Drones are the wedge.

Market

The market is $65B today and $1.7T by 2035.

Commercial drone hardware + services, all segments. Conservative analysts project mid-teens CAGR; aggressive ones project ~30%. DRONA sits one layer up — the design and operations surface every builder and operator touches before they buy a part. Capture is measured in bps of hardware GMV, not headcount.

$65B
Global drone market · 2025
$117B
Forecast · 2030 · 12.5% CAGR (Mordor, Grand View)
$1.4–1.7T
Forecast · 2035 (Precedence, Skyquest)
$22.5B → $51.7B
Commercial UAV narrow slice · 2026 → 2030 · 23% CAGR (R&M)

Where DRONA collects rent

We do not sell drones. We sell the design layer — subscription per seat for prosumers, enterprise tier for operators, take rate on the BOM for marketplace flow. Three revenue surfaces, one product.

Prosumer / SMB seats$29–79 / mo
Farmers, inspectors, filmmakers, university teams, hobbyists who want a real answer. Analog: Figma seat.
SAM$3.2B
Enterprise fleet tier$60k–400k ACV
Utility operators, agri-ops, public-safety departments, defense primes. SSO, audit trails, Blue UAS filter, private catalog. Analog: Palantir Foundry seat.
SAM$8.4B
Marketplace take rate3–6% of BOM
Every exported .vdx hits a sourcing surface. Retailer partner, not competitor — we route orders, they fulfil. Analog: Shopify payments.
SAM$4–9B by 2030
Why now · why us

Three curves are crossing in 2026. One team is already at the intersection.

01

LLM tool-use got reliable 18 months ago.

Structured tool-calling crossed the threshold where a language model can drive a typed physics engine without hallucinating. That is the wedge — chat as interface, physics as authority. Before 2024 this did not work. In 2026 it does.

02

Commercial drone use overtook consumer.

Solar inspection grew from 200 MW/yr in 2019 to 28 GW in 2025. Agriculture went from hobbyist to $4.2B. BVLOS rules finally shipped in five major jurisdictions. The buyer exists, is spending, and does not want a hobbyist workflow.

03

No incumbent owns the design layer.

DJI owns the hardware. Skydio owns the autonomy. Auterion owns the firmware. Nobody owns the conversation between "what do you need done" and "here is the machine." The layer is unclaimed, and the window is roughly 24 months.

Why this team

Engineering-heavyFirmware hackers, computer-graphics engineers, applied-AI researchers. Small. No growth marketers yet.
Domain-embeddedWe fly the drones we design. We ride along on solar runs. We talk to farmers and marine biologists weekly.
Physics-firstThe compatibility engine ships outside the model, exposed as typed tools. Architectural discipline others skip.
Open artifactEvery design exports to a portable .vdx spec users own. No lock-in is our moat, not our risk.
Industries · dollar impact

Eight sectors. Each one a standalone unicorn. Each one a DRONA row.

Below: per-sector GMV that DRONA designs will touch by 2030, conservatively modelled at 0.5–2% of annual services spend. The full sheet is in the data room.

Agriculture
Precision-ag drone services — spraying, scouting, livestock locate. DRONA designs the rig; the rig runs the field.
$12B
Solar inspection
Annual thermal + defect scanning · 1.2 TW installed · 8% inspected annually · $0.40/panel fully loaded.
$3.8B
Search & rescue
Equipped-unit market by 2030 · ~$180M in saved lives per annum at current response-time deltas.
$900M
Delivery & logistics
Last-mile parcel + medical · by 2030. DRONA designs the vehicle, not the network.
$28B
Film & cinema
Drone production services · broadcast tether, FPV action, cinema A-cam. Long tail, high margin.
$1.1B
Infrastructure inspection
Powerline, wind-turbine blade-follow, pipeline, bridge. Mandatory in four jurisdictions.
$9.6B
Mapping & survey
Cadastral PPK, forest LiDAR, mine volumetric. Most regulated drone category; stickiest retention.
$5.4B
Underwater · marine
Reef survey, aquaculture, subsea pipeline AUV. Q4 2026 form factor. Pure greenfield.
$2.1B

Every number above is the denominator. Our numerator — 0.5% to 2% of it — is what we are raising to capture. The math is in the data room; the thesis here is simple: DRONA does not pick a vertical — it picks all of them.

What the round funds

Twelve months. Four milestones. One standard.

45%
Engineering headcount
Six hires · two firmware, one graphics, two applied-AI, one simulation. Small by design.
20%
Catalog & physics coverage
From 834 SKUs to 6,000. From multirotor + fixed-wing to full seven-form-factor parity. Verified weekly.
15%
Enterprise tier GA
SSO · SAML · SCIM · audit trail · Blue UAS filter · private catalog. First five design-partners signed.
12%
Community & field work
Drone meetups · DBF sponsorship · field days with farmers and solar operators. Content follows users.
8%
Infra & compliance
Vector DB scale · export compliance · regional regulatory profiles · SOC-2 Type I.

Twelve-month milestones

Q2 2026
Ship Fixed-wing end-to-end · industry × autonomy matrix · photo-in-chat GA
shipping now
Q3 2026
Ship VTOL · tilt-rotor · quadplane · transition-envelope math
dev complete
Q4 2026
Ship Collective-pitch heli · ROV · airship · public fork network
planning
Q1 2027
Ship Enterprise tier GA · five paying design-partners · $500k ARR target
gated
Traction · live

Private beta. Real numbers. Updated weekly.

{{WAITLIST}}
Waitlist signups
{{SESSIONS}}
Design sessions completed
{{BUILDS}}
.vdx builds exported
834
Verified parts in catalog
{{COUNTRIES}}
Countries with active users
{{NPS}}
NPS · first 100 users
{{SHIPS}}
Ships · last 12 weeks
{{D30}}
Day-30 retention · beta cohort
Live · last updated {{UPDATED}} Source · production telemetry
Who is already in

Existing backers, advisors, and operators alongside this round.

Lead · {{LEAD_FUND}}{{LEAD_PARTNER}} · board seat
Seed angels · {{ANGEL_COUNT}}Drone operators · robotics founders · one ex-DJI exec
Advisor · hardware{{ADVISOR_HW}} · ex-Skydio · shipped 3 autonomy stacks
Advisor · ag-tech{{ADVISOR_AG}} · runs a 12,000 ha operation · first customer
Advisor · defense{{ADVISOR_DEF}} · Blue UAS standards council
Advisor · research{{ADVISOR_RES}} · published compat-engine paper with us

Names shared under NDA at first call.

Request allocation

This is a whitelist, not a form.

We read every submission. We reply within 72 hours. We keep the round tight — four answers decide whether we set up a call.

You'll get a magic link to a private data room with deeper financials, references, and a 30-min call slot.

Got it. We read these personally.

Check your inbox for a confirmation and, if there's a fit, a calendar link within 72 hours. If you don't hear back, it is not a no — just a full queue. Nudge founders@drona.design.

FAQ

Five questions every investor asks us first.

Isn't drone design a niche hobbyist market?

It was, in 2016. In 2026, 74% of drone GMV is commercial — agriculture, inspection, mapping, delivery, public safety. The hobbyist is the entry gateway, not the ceiling. Our first enterprise contract will be larger than our first 200 prosumer seats combined.

What stops DJI or Skydio from building this?

They sell hardware. We sell the layer that routes orders to every hardware vendor. Our incentive is multi-vendor; theirs is single-vendor. A neutral design layer is orthogonal to their P&L — the same reason Figma wasn't built by Apple.

How do you prevent AI hallucination of part numbers?

The model never speaks part numbers directly. It calls a typed tool against a verified catalog. Every returned SKU is tagged grounded, estimated, or conceptual. The compat engine is deterministic ECMAScript — the same module runs server and client. The architectural choice is permanent.

Weapons systems?

Hard no. It is in the system prompt, in the tool schemas, and in the manifesto. We will not help design weapons, loitering munitions, lethal autonomous systems, detection-evasion rigs for criminal use, or mass-targeting configurations. We are happy to keep that line bright even when it costs us customers.

What happens if the round doesn't close?

We ship anyway, slower. We are currently default-alive at current burn through Q3 2026 with existing revenue. The round accelerates the enterprise tier and form-factor coverage — it does not keep the lights on.