
Applications are being sought from Irish-based AI start-ups to participate in the 2026 AI Ecosystem Accelerator programme at University College Dublin (UCD). The accelerator is delivered by NovaUCD and CeADAR as part of CeADAR’s European Digital Innovation Hub for AI programme which supports Irish start-ups and SMEs innovating with AI.
Now in its third year this 6-month accelerator is designed to support Ireland’s most promising and disruptive AI-first start-ups to accelerate commercial traction for global markets and become investment-ready scalable businesses, with a particular focus on start-ups actively raising and aiming to close a funding round in 2026.
The AI Ecosystem Accelerator programme provides a structured pathway for founders who have moved beyond early ideation and are ready to convert validated momentum into scalable growth and investor readiness.
The programme also delivers a combination of commercial, fundraising and technical supports, including guidance from CeADAR’s AI experts to strengthen AI implementation and readiness for scale. This includes support on data strategy, model evaluation, AI deployment readiness, and ensuring the underlying AI is robust, reliable and fit for real-world use.
A total of twenty (20) start-ups have completed the programme in the last two years. The 2026 programme, which starts in April, will culminate with an Investor Showcase event, connecting founders with investors and strategic partners at the point they are ready to progress and accelerate their next stage of growth.
Targeted supports for programme participants also include experienced commercial mentors and practical workshops focused on positioning, value proposition refinement, scaling and investor readiness. Participants will also benefit from access to collaborative co-working space at NovaUCD, with opportunities to connect with successful entrepreneurs, start-ups, corporates, investors and access to skilled interns.
Applications are open via https://www.ucd.ie/innovation/start-ups/aiecosystemaccelerator/






