In the press, on stage, on record.

DATAmundi wins a Globee Innovation Award for AIDA Agents

AIDA Agents, the multi-agent orchestration platform whose research team and implementation I led at DATAmundi, won a Globee Innovation Award in the Best Multi-Agent System Orchestration Platform category. The award recognizes technology built to manage complex AI data operations and multilingual content workflows. Before that, the Business Intelligence Group named AIDA Agents a finalist in the 2026 Artificial Intelligence Excellence Awards.

“We’re delighted to see AIDA recognized by the 2026 Globee Innovation Awards. We can call the current AI phase as the Enterprise Agentic Era: it is defined not simply by increasingly powerful individual models, but by how intelligently we orchestrate specialist agents around them in enterprise applications. We anticipated this phase over a year ago at research level, when we targeted agentic orchestration as the new, potentially, key enabler to deliver significant quality and efficiency improvements in both localization and data services. AIDA Agents demonstrates how well that approach works while building evaluation, governance, and human expertise directly into the workflow.”

DATAmundi Tests AIDA Agents on General and Terminology-Constrained Translation

Slator covered the benchmark results behind AIDA Agents, the multi-agent platform whose research team and implementation I led at DATAmundi: translator, rater, post-editor and re-rater agents orchestrated with no model fine-tuning, reported in two peer-reviewed papers (EAMT 2026 and ACL 2026). Best on 10 of 11 WMT24++ language pairs; 70–98% of segments publication-ready on an industrial benchmark; three configurations — AIDA LLM, AIDA MultiLLM and AIDA NMT — to trade quality against cost.

Presenting at ACL 2026 — the top conference in computational linguistics

On July 7 I presented “Multi-Agent Orchestration for Terminology-Constrained Machine Translation in Industrial Localization” in the ACL 2026 Industry Track in San Diego, with DATAmundi as a Bronze sponsor of the conference. The paper reports 99.4% terminology accuracy and the best ChrF2++ on the WMT25 terminology task, from a system running in production.

It has since been cited as related work by researchers at JPMorgan AI Research (Findings of EMNLP 2026): “reasoning-enabled review can be effective for segment-level terminology-constrained MT”; and at Amazon (WMT 2026): “structured agent coordination can reliably enforce domain glossaries at scale.”

“Beyond LLMs: What’s Next in the AI Tech Stack for Multilingual Software?”

Invited panelist at Adobe’s annual globalization summit. I spoke about agentic AI, with a focus on multi-agent collaboration workflows applied to machine translation and multilingual data-quality pipelines, and about the work we do at DATAmundi in this area, building on the latest research.

On the panel with Nitin Aggarwal (Sr Director, Generative AI, Microsoft), Adam Bittlingmayer (Co-Founder & CEO, ModelFront) and Kirti Vashee (Language Technology Evangelist, Translated), moderated by Renato Beninatto (Chief Advisor, Nimdzi Insights). The keynotes came from Spotify, Microsoft, Uber, Google and Adobe; the speakers received the summit’s speaker award.

Meeting Łukasz Kaiser, co-author of “Attention Is All You Need”

Łukasz Kaiser, one of the inventors of the Transformer, opened the conference with a keynote on the path from RNNs to LLMs and now to reasoning models, and on how they translate creative, nuanced content such as poetry. We talked about the latest research trends in language models, the impact of reasoning, and smaller models in industrial applications, and I told him what we are building at DATAmundi.

The Transformer paper was written with machine translation in mind, and went on to change almost every other field.

Summa Linguae Technologies Welcomes Emanuele Di Rosa as Chief Technology Officer

The appointment announcement: almost two decades leading AI and software projects for global brands including Stellantis, Samsung and Adidas, and expertise in AI, language processing and cognitive technologies going back to pioneering chatbot work in 2004. The DATAmundi leadership page adds the mandate: modernizing localization workflows with LLM technologies and building a robust platform for multilingual data services.

“Summa Linguae is in a transformative phase, strengthening its offering with AI technologies at the core. I see a significant opportunity to enhance our service delivery while maintaining our commitment to quality.” … “The goal is not just to work faster but to make better decisions and deliver higher-quality results through AI-enhanced capabilities.”

Innovation Award from Pitney Bowes, for EVA

Pitney Bowes gave App2Check its Innovation Award for the AI and natural-language-understanding innovation we brought into EVA, our platform for cognitive chatbots. EVA’s integration with Pitney Bowes EngageOne Converse strengthened both platforms and made the chatbots smarter, adding value for our customers.

“EVA by App2Check: Build a Cognitive Chatbot With No Technical Skills and Maintain its Knowledge Base as a CMS”

Speaker at the world’s largest bot conference, presenting EVA — the no-code cognitive-chatbot platform built on our patented knowledge-base extraction — jointly with Pitney Bowes, whose EngageOne Converse platform integrated EVA. The organizers’ own photo caption: “Emanuele Di Rosa from App2Check gave a fantastic pitch…”

“CX Insights Discovery and Emotion Tracking”

Speaker at CX Emotion 2019 in London, as CTO of App2Check, on discovering customer-experience insights and tracking emotions with App2Check’s Voice-of-the-Customer analytics. The organizers published the deck.

App2Check at the Pitney Bowes EMEA Conference 2018

As a Pitney Bowes business partner, we presented the integration of the EVA chatbot with the EngageOne Converse customer-engagement platform — AI that builds a knowledge base in seconds and starts answering customers with a high level of accuracy. The partnership later became Precisely’s worldwide partnership after the 2020 rebrand.

Shortlisted in Agos’ “Worldwide AI Software Vendor Selection”

On stage at Agos’ Artificial Intelligence Day, I presented our email-bot with churn prediction, integrated with the company CRM: machine-learning text analysis that flags customers at high risk of churn and proposes offers to keep them. After a technical screening of more than 300 companies worldwide, App2Check was one of five shortlisted in the Customer Interaction track, and in 2018 we joined the Agos Ducato Innovation Lab.

“Comparing Sentiment Engine Performance on Reviews and Tweets”

My talk at the Sentiment Analysis Symposium 2017 in New York, the industry gathering on sentiment, emotion and intent analytics: the state of the art in sentiment-engine performance reached by both research and industrial tools, on reviews and on tweets.