December 29, 2025 1 min to read

Happy New Year 2026

Do not dwell in the past, do not dream of the future, concentrate the mind on the present moment. - Buddha

Year 2026 - A New Start, rather a restart!

Restart 2026 to make difference, lasting impact with another 365 days at your disposal.

In 2026, may your creativity scale like parameters in a neural net, your insights converge like reasoning paths, and your actions embody the initiative of agentic intelligence.

Happy New Year 2026! May this year bring you prosperity, joy, and endless opportunities to explore and innovate. Together, let’s write another remarkable chapter in our journey of AI excellence.

Time to reflect back on 2025 and look forth for 2026 with a new perspective.

Agentic AI has been talk of the year 2025, is it here to stay?

Agentic AI dominated 2025 discussions and deployments, transitioning from hype to production tools in enterprises, with coding agents achieving breakthrough SWE-Bench scores around 60% resolved rate.

2025 Agentic AI Highlights

Agentic systems evolved into multi-agent collaborations and self-healing setups, resolving up to 80% of customer service issues autonomously per Gartner forecasts realized in pilots. Vertical-specific agents for industries like healthcare and finance gained traction, while human-in-the-loop governance addressed risks in scaling autonomy. Coding agents like Claude Code and GPT-5 variants led, with SWE-rebench showing top models at 62.1% resolved (Claude Code) and 74.5% pass@5, far surpassing 2024 baselines.

Coding Agents Rock SWE-Bench

Frontier models in late 2025 hit 60%+ on SWE-rebench (e.g., Claude Sonnet 4.5 at 60.9%, GPT-5.2 at 61.3%), using efficient token caching and reasoning modes. SWE-Bench Pro remains tougher at ~45% for leaders like Claude Opus 4.5, highlighting gaps in complex, multi-file enterprise tasks. Open-weight standouts like DeepSeek-V3.2 (46.4%) and MiniMax M2 (39.6%) offer cost-effective alternatives at $0.10–$0.56 per problem.

Is Agentic AI Here to Stay?

Yes—2026 predictions emphasize mainstream adoption with predictive agents, multi-agent squads in ITSM, and 2-3x efficiency gains, per Microsoft and Deloitte analyses. Challenges persist in reliability (e.g., 60% developer sentiment dip) and governance, but production successes in SDLC phases ensure persistence. Expect agentic AI to redefine workflows, with coding agents enabling solo devs to match teams via tools like Cursor and GitHub Copilot.

…..

If you are interested in Citizen Development, refer to this book outline here on Empower Innovation: A Guide to Citizen Development in Microsoft 365

Now, available on