Eight companies have joined the newest cohort of the Tehnopol AI Accelerator in Estonia, working on AI-based products spanning education, compliance, procurement, cloud documentation and second-hand retail.
The Tehnopol AI Accelerator, which is part of an AI development programme backed by the Estonian Ministry of Economic Affairs and Communications, supports technology companies in bringing AI-based products to market through business model development, customer validation and mentoring. At the end of the programme, a jury can award up to three startups a development grant of 10,000 euros each.
Among the eight new participants, DoksOps automates the generation of audit-ready documentation from AWS cloud infrastructure, covering standards such as ISO 27001, SOC 2 and NIS2, and is already available on AWS Marketplace. Aelethion offers a product called Athena that gives organisations visibility and control over how AI tools handle data and take actions. PunanePastakas, which has won the Estonian President’s Education Hackathon and is being tested by around 40 teachers in Estonia, uses AI to help STEM teachers assess handwritten student work and generate feedback drafts.
AveoTech builds AI agents to automate quotation, pricing and procurement workflows from initial enquiry through to a signed document. Mergo connects customer feedback from support requests, reviews and surveys with product usage data to help teams prioritise which problems to address. Auctologue analyses photographs of antiques, collectables and second-hand goods to produce detailed, search-optimised product listings for auction houses, e-commerce platforms and vintage retailers.
Atlaswiki structures company knowledge into attributable facts, identifies contradictions and keeps information current to improve the reliability of AI systems that draw on internal data. Kimova AI is building an AI-powered compliance audit tool designed to handle routine audit tasks while keeping substantive decisions under human control.
Source: Tehnopol. Read the original (published 2026-11-11).
