QED Software
Warsaw, Poland
About
QED Software is a Warsaw AI consultancy building language-model and machine-learning systems, including explainable AI tooling, for commercial clients.
Who They Work With
Data-rich firms and startups with a detection, vision or forecasting problem suited to a research-led team.
Based on 7 client engagements with a source URL · researched 27 Sept 2026
Our Take
QED Software's strongest evidence sits in its client quotes rather than its case studies: DeepSeas' CTO credits models drawn from a contest QED ran on its KnowledgePit.ai platform with 25% more efficient alert analysis in his security operations team, and FitFood's CEO describes a demand-forecasting system integrated into the company's logistics.
The six published case studies name the client but rarely the method—the DeepSeas study describes no model and no result, the Vertis study claims a strategic AI deployment without describing a system, and the ONSTAGE.ai figure of up to 60% lower production cost is sourced to industry data rather than measured.
The clearer technical stories are ARAHUB's Blindsight image encoding, which kept camera-equipped ad screens running under GDPR, and a satellite crop-segmentation model delivered as a QGIS plug-in.
The weaker fit is buyers who need published accuracy figures or a production reference they can check before signing—the homepage's "Stop Prototyping. Start Deploying." runs ahead of what the case studies document.
Key strengths
- Founder Prof. Dominik Ślęzak and two named leads, Jacek Puczniewski and Piotr Biczyk PhD, appear on the homepage; the EU projects page lists grant application numbers and project values
- Three client executives are quoted by name and title—DeepSeas' CTO, FitFood's CEO and Tricholab's President—and the DeepSeas quote carries a figure: 25% more efficient alert analysis
- The ARAHUB study explains its privacy mechanism: frames encoded so a person cannot be recognized while age and gender still can, streamed without storing the original image
Good to know
- Case studies publish almost no stack, accuracy or volume figures—ask for the evaluation results and production setup behind the DeepSeas or FitFood work before scoping
- Every case study carries a November 2025 date, which hides when the work was done—ask which engagements are LLM-era and which predate them
- The homepage counts 40 AI/ML experts "in the network", not on staff—ask which employed engineers and which named leads would work on your project
Reviewed by Gabor Kiss
Founder & Curator, AIAgencies.eu · 23 Sept 2026
Based on the agency's publicly available portfolio. Own this agency? Claim your profile to provide additional context or request a review update.
Notable Clients
Every engagement documented in a published case study
Documented engagements: DeepSeas, ARAHUB, KidiHub, ONSTAGE.ai, Vertis, FitFood, and Tricholab. All 7 carry a source URL on file.
Industries
Get Matched with Agencies
RecommendedTell us your needs and we'll connect you with 3 vetted agencies—including QED Software if it's a fit.
Get Matched →Curator shortlist within 3 business days
Visit QED Software's website →Own this agency? Claim your profile to verify info & receive leads directly.
