● Filing No. 001 — May 2026
Accelerating food
abundance with robotics.
Universal Robots
NVIDIA Inception
MassRobotics
REV Hardware Accelerator
Cornell · Blackstone LaunchPad
Hilton LOI
Mazra Restaurant Group
Universal Robots
NVIDIA Inception
MassRobotics
REV Hardware Accelerator
Cornell · Blackstone LaunchPad
Hilton LOI
Mazra Restaurant Group
75+
Customer interviews
3
LOIs secured
2
LOIs signed
~90%
AI accuracy validated
§ 01 · Velocity
Eight months.
Four prototypes.
Deployment in August.
SEP 2025
Idea + Alpha 1
$60K raised. 30 customer interviews. First LOIs signed within 30 days of incorporation.
OCT 2025
Proto 2 — full rebuild
Hardware and AI stack rewritten from scratch after customer feedback. Second iteration validated multi-modal vision pipeline.
NOV 2025
MassRobotics admission
Accepted into the Boston robotics ecosystem. Began full-time relocation of operations.
FEB 2026
UR + NVIDIA + REV
Universal Robots loan agreement. NVIDIA Inception membership. Accepted to REV Hardware Accelerator AgTech track.
APR 2026
Proto v4 — CAD complete
Production-intent mechanical and electrical design finalized. Custom end-effectors, feeding mechanism, and cutting system locked for fabrication.
MAY 2026
Full-time at MassRobotics
Founding team going full-time on Ficio. Living together at C-House in Boston. Operating out of MassRobotics — eating, sleeping, and shipping in the same building.
JUN — JUL 2026
Proto v4 testing
Three production-intent units fabricated through REV. End-to-end testing on customer ingredients. AI stack validated against real kitchen variance.
AUG 2026
First customer deployment
FicioPrep deployed in commercial kitchens by end of August. First revenue. First fleet data feeding the AI loop.
§ 01b · In the lab
Watch the build.
§ 02 · Vision
Where this goes.
Kitchens that operate just-in-time — where every ingredient is prepped seconds before it's needed, at Michelin-grade quality, with nanometer-precision consistency, and an optimization layer that matches idle capacity to live demand across a continent.
01 / Wedge
FicioPrep
Drop-in ingredient prep with multi-modal AI quality control. ~90% classification accuracy on potatoes alone.
02 / Expansion
FicioChef
Generalized cooking — searing, stirring, frying. The same AI brain, different end-effector.
03 / Platform
Kitchen OS
Idle kitchens become a 24/7 production network. Off-hours capacity rented to ghost brands and delivery platforms.
§ 03 · Timing
Why now.
01
VLA models crossed the threshold
Vision-language-action models can now generalize across food objects without per-SKU training. Two years ago this was a research demo. Today it runs on a Jetson at the edge.
02
Kitchen labor structurally broken
75% turnover. $218B in food waste annually. Operators are no longer asking if they should automate — they're asking who can deploy first.
03
Hardware finally affordable
Collaborative arms at commodity prices. Custom end-effectors at $7.5K COGS. We deliver the same outcome as a $1M install for $1,500/month.
§ 04 · Demand
Who's asking.
Hilton
Global hospitality
⎯ LOI signed
Mazra
Restaurant group
⎯ LOI signed
Calla / HGI
Monterey hotel
⎯ Pilot signed
Bakersfield
Commissary kitchen
⎯ Pilot signed
Marriott
In conversation
⎯ Active outreach
UCLA Dining
University foodservice
⎯ Active outreach
Cornell Dining
University foodservice
⎯ Active outreach
Boston Airport
Commissary network
⎯ Intro confirmed
§ 05 · Operators
Built by people who've shipped this before.
Youhanna Meleka
Co-founder · CEO
Tesla Top Talent 2024. Architected end-to-end manufacturing optimization that saved $43M. Cornell M.Eng. — Photonics & Quantum Engineering. 7+ years manufacturing.
David Farag
Co-founder · AI / EE
Google at 18. Reduced ARCore app size by 20%. Built three functional robots and won UCLA robotics competition as a freshman. Self-taught programmer at 12.
Ahmed Rashad
Co-founder · Mech / Controls
Led full-stack design and instrumentation for $100M+ industrial projects. Currently leads a 45-person technical team through commissioning operations.
Daniel Bishara
Founding · EE / SWE
Tesla — optimized Optimus motor control, doubling throughput. Apple iPhone failure analysis and power integrity. Waterloo.
Why us.
Manufacturing depth from Tesla. AI depth from Google and Apple. Hardware shipped, not whitepapered. Maniacal urgency to end hunger.
※ Verified · Confidential
How we operate.
Ficio is an AI-first company. Every customer call, research paper, and decision is logged, embedded, and queryable. Our institutional memory compounds while competitors forget.
01
Company Brain
A unified RAG layer over every meeting, document, and Slack thread. Any team member can ask "what did we decide on UR IP terms" and get a sourced answer in seconds.
Internal · Live
02
CREST
Our framework for converting cutting-edge research into shipped product. We read papers, extract methods, prototype on internal datasets, and deploy what works to FicioPrep — weekly cadence.
Research → Production
03
AI Sales
Every customer call is transcribed, analyzed, and surfaced as a dashboard. Who we talked to, what we learned, what objections recurred, and how the next call should be calibrated.
Outreach Engine
04
AI Supply Chain
Bill-of-materials sourcing, vendor outreach, and lead-time tracking automated end to end. Manufacturing depth turned into a software loop.
Continuous Cost Advantage
Data Room
Financial model, cap table & hardware deep dive.
Full data room available to verified investors on request.