This Week in AI & Enterprise Innovation
Welcome to this week’s edition of NexaQuanta’s AI & Enterprise Innovation Newsletter. As organisations rapidly adopt automation, agentic AI, and data-driven decision systems, the landscape continues to evolve at an unprecedented pace.
- This week, Amazon expanded its Nova AI suite to help companies build custom models and automate operations end-to-end.
- IBM and Pearson announced a significant partnership to address global skills gaps with new AI-powered learning tools.
- Google launched an upgraded Gemini Deep Research agent to strengthen enterprise analysis and reduce hallucinations in long-running tasks.
- OpenAI released GPT-5.2, its most advanced model yet for professional work and agentic workflows.
- And at Ignite 2025, Microsoft introduced five significant AI and cloud upgrades across Foundry, Azure HorizonDB, Copilot, and next-gen hardware.
Amazon Unveils Nova AI Suite to Help Businesses Build Custom Models and Automate Operations
Enhanced Access to Enterprise-Grade AI
Amazon has expanded its Nova portfolio with four new AI models designed to bring advanced capabilities to small and mid-sized businesses. The Nova 2 lineup offers improvements in reasoning, multimodal processing, conversational AI, code generation, and agentic tasks.
Custom Model Training Through Nova Forge
A significant addition to the portfolio is Nova Forge, a service that allows organisations to build customised models using an “open training” method. Businesses can integrate their proprietary data early in the training process, enabling more relevant and specialised AI applications.
Automation Gains with Nova Act
Amazon also introduced Nova Act, a tool focused on UI-based task automation. Early customers report up to 90% reliability in browser workflow automation. For small businesses, this can translate into reduced manual effort and more streamlined digital operations.
Considerations for Small Businesses
While the Nova expansion opens doors for broader AI adoption, businesses may still face challenges related to technical readiness, data security, and implementation costs. Integrating proprietary data into custom models requires strong governance practices, and some teams may need additional training or specialised talent.
Want to read more about the news? Click here!
IBM and Pearson Collaborate to Build New AI-Powered Learning Tools for Organisations
Responding to the Skills Gap
IBM and Pearson have announced a global partnership to develop new AI-powered learning products for enterprises, public organisations, and educational institutions. The collaboration comes as Pearson research indicates that skills mismatches and slow career transitions could cost the U.S. economy $1.1 trillion in annual lost earnings.
New Learning Solutions Built on watsonx
Using IBM’s watsonx Orchestrate and watsonx Governance, the two companies will build personalised learning tools designed to support large-scale workforce development. IBM will also help Pearson create a custom AI learning platform, similar to IBM Consulting Advantage, that blends human expertise with AI assistants and agents.
Strengthening Workforce Transformation
As part of the partnership, Pearson will serve as IBM’s primary strategic partner for upskilling initiatives. IBM’s clients—and its 270,000 employees—will gain access to Pearson solutions such as Credly for digital credentials, Faethm for workforce planning, and Pearson Professional Assessments, which delivers IBM certification exams worldwide.
Strategic Alignment
The partnership supports Pearson’s strategy to build long-term relationships with key partners and co-create new market opportunities. Both firms position this collaboration as a step toward helping businesses and individuals adapt more quickly to the demands of the AI era.
Click here to read more about this news.
Google Launches New Gemini Deep Research Agent as Competition Heats Up with OpenAI’s GPT-5.2
A Major Upgrade for Enterprise Research
Google has introduced a redesigned version of its Gemini Deep Research agent, built on the company’s flagship Gemini 3 Pro model. Unlike earlier iterations that focused mainly on generating research reports, the new version allows developers to embed Google’s advanced research capabilities directly into their own applications through the new Interactions API.
Built for High-Precision, Large-Context Analysis
The updated Deep Research agent can process large volumes of contextual data and synthesise information across domains, including due diligence, scientific analysis, and drug toxicity research.
Google plans to integrate the agent into Google Search, Google Finance, the Gemini app, and NotebookLM, signalling a shift toward environments where AI assistants—not users—perform deep information gathering.
Tackling Hallucinations in Agentic AI
Google positions Gemini 3 Pro as its “most factual” model, optimised to reduce hallucinations during long-running tasks. To validate performance, the company released a new benchmark, DeepSearchQA, for testing multi-step information-seeking tasks and open-sourced it to the research community.
Competitive Timing with OpenAI
Google’s announcement arrived just as OpenAI released GPT-5.2, codenamed Garlic, a model OpenAI claims outperforms competitors—including Google—across several benchmarks. The timing underscores the intensifying rivalry between the two companies as both push for leadership in enterprise-grade agentic AI.
Click here to read more about the news.
OpenAI Introduces GPT-5.2, Its Most Advanced Model for Professional Work and Long-Running AI Agents
Built for Enterprise Productivity
OpenAI has launched GPT-5.2, its most capable model series to date, designed to support professional knowledge work and complex, multi-step agentic tasks.
The company says ChatGPT Enterprise users already save 40–60 minutes per day, and GPT-5.2 aims to increase that value by strengthening its abilities in spreadsheets, presentations, code generation, long-context understanding, tool use, and image interpretation.
Strong Performance Across Industry Benchmarks
GPT-5.2 sets new records on multiple evaluations. On GDPval—a benchmark covering knowledge tasks across 44 occupations—the model beats or ties industry professionals in 70.9% of comparisons. It also delivers improved results in software engineering, scientific reasoning, abstract problem-solving, mathematics, and long-context analysis.
Expanded Tools and Enterprise Rollout
GPT-5.2 is now available in ChatGPT under the Instant, Thinking, and Pro variants for paid plans. Developers can access it immediately through the API.
OpenAI notes that the new Thinking model is particularly effective for real-world outputs such as sales presentations, financial models, staffing plans, and operational documents—producing results significantly faster and at a fraction of the cost of traditional approaches.
Want to read more? Click here!
Microsoft Unveils Five Key AI & Cloud Innovations at Ignite 2025
Microsoft announced major upgrades across AI, data, and cloud infrastructure at Ignite 2025, reinforcing its push toward agentic, enterprise-ready AI.
- Claude Models on Microsoft Foundry
Anthropic’s Claude Sonnet and Opus models are now available on Foundry, giving enterprises broader options for frontier models on Azure.
- Foundry IQ & Fabric IQ Launched
New IQ tools enable deeper semantic understanding and policy-aware search, simplifying the development of agentic enterprise applications.
- Azure HorizonDB Introduced
A new Postgres-compatible, AI-optimised database delivering higher performance, vector indexing, and seamless integration with Foundry.
- Azure Copilot Gains Agentic Operations
Expanded agents now support end-to-end cloud operations—from migration to deployment and optimisation—accelerating modernisation.
- Next-Gen Azure Hardware
Microsoft showcased new custom silicon and NVIDIA Blackwell Ultra infrastructure to support high-performance AI workloads.
Stay Connected
Thank you for reading this week’s edition. If you want to stay ahead of fast-moving AI trends and receive curated insights every week, we encourage you to subscribe to our newsletter. Join a growing community of leaders who rely on NexaQuanta for clear, business-focused updates on the technologies shaping tomorrow’s enterprises.