Ankur Kumar

5️⃣ Nebius → Tavily bringing real‑time “agentic search” infrastructure into its AI cloud platform

3️⃣ Koyeb acquired by Mistral AI to build the next-generation of cloud infra for AI
4️⃣ IBM → Confluent acquisition framed as building a smart data platform for connecting and governing data for AI applications and agents

As anticipated, Q4 ‘25 & Q1 ‘26 witnessed consolidation of AI-first companies 👇
1️⃣ Manus acquisition by Meta, accelerating its development of autonomous coding AI agents
2️⃣ OpenAI hires OpenClaw creator Peter (beating much anticipated Anthropic offer), consolidation of personal AI Agents

#AgenticAI

Liked the context management aspect of the repository 👇

Attention is all we need 👍🏻

3️⃣ Adaptive thinking is another useful tool- Claude evaluates the complexity of each request and decides whether and how much to think
4️⃣ Automated Context Compaction: automatically summarizes and replaces older context when the conversation approaches a configurable threshold

1️⃣ 1M token with Agent Teams applying the Planner-Orchestration-Specialized Team of Agents pattern baked in
2️⃣ Autonomous Multitasking, taking a part of burden away from downstream systems

www-cdn.anthropic.com/14e4fb01875d...

With Claude Opus 4.6, Anthropic has raised the bar further particularly for coding agents with shift-left mindset such as (they don’t release any internal architecture details such as whether it uses mixture-of-experts)👇

www.anthropic.com/news/claude-...

Shared consolidated perspective on key technology trends to watch for in 2026.
Read detailed article here 👇
medium.com/vedcraft/top...

What Agentic AI standard you are using or planning to?
1️⃣ AGENTS.md
2️⃣ agentskills.io
3️⃣ modelcontextprotocol.io
4️⃣ a2a-protocol.org
5️⃣ mcpui.dev

Any other - please comment.

goose your open source AI agent, automating engineering tasks seamlessly

3️⃣ Block provided Goose(block.github.io/goose), an open source local-first AI framework

This will encourage a suite of Open source Agentic AI initiatives to be part of common set of standards and technologies.

#AgenticAI #GenAI #LinuxFoundation
#MCP #AGENTS

AGENTS.md AGENTS.md is a simple, open format for guiding coding agents. Think of it as a README for agents.

2️⃣OpenAI offered the AGENTS.md specification that gives AI coding agents consistent, project-specific knowledge

What is the Model Context Protocol (MCP)? - Model Context Protocol

1️⃣ Anthropic contributed Model Context Protocol (modelcontextprotocol.io), a universal open standard for connecting LLM apps with contextual data

That’s a great news for Open Source towards establishing common standards and technologies as a community - The Linux Foundation launched Agentic AI Foundation (aaif.io) joined by Anthropic, OpenAI, Google, Microsoft, AWS, Cloudflare, Bloomberg, and Block 👇

Agentic AI Gateway: The Proven Architecture Pattern for Enterprise GenAI Security and Governance Agentic AI Gateway: The Proven Architecture Pattern for Enterprise GenAI Security and Governance With the rise of Agentic AI, establishing enterprise-wide architecture building blocks has emerged as …

Agentic AI Gateway: The Proven Architecture Pattern for Enterprise GenAI Security and Governance. Read more for key insights 👇

#AgenticAI #GenAI

medium.com/vedcraft/age...

🧮 Information retrieval from internal sources (unstructured)
🧭 Information retrieval from internal sources (structured)
📚Vector storage & database deployment patterns
🔏 Pre and Post filtering of content as per security guardrails

📐Reranking of the information retrieval
📌 Compression of the information retrieval
📝 Structured information retrieval optimization
🔎 Information retrieval from public sources

AI Bits - While building Agentic Apps, the Retrieval of relevant information from structured or unstructured sources plays the pivotal role, and there are many design considerations to be made 🧵

#GenAI #AgenticAI

An illustrative maturity model (image generated with Nano Banana)

Stage 3 (Run): Build or leverage bespoke and enterprise-specific autonomous AI agents scaling organizational efficiency and achieving new possibilities

Stage 2 (Walk): Build enterprise or product-specific AI Assistants for individuals & organization productivity (e.g. Google Cloud Assist, Amazon Q, Microsoft Copilot, Salesforce Einstein, ServiceNow Now Assist, Oracle AI Agents, IBM watsonx Assistant, Adobe Firefly Assistant, Workday AI, Zoom AI)

Stage 1 (Crawl): Building the foundation layer and Agentic AI platform for building AI assistants and Autonomous Agents

AI Bits#7 - That’s the vision getting formulated across the leading product companies and enterprises 🧵

✔️Berkeley Function/Tool Calling: gorilla.cs.berkeley.edu/leaderboard....
✔️LMArena: lmarena.ai/leaderboard
✔️SWE Bench: https://

2️⃣ Apply Auto-Reasoning and Auto-Selection of Models using solutions such as Semantic Routing instead of static binding
3️⃣ Reference existing industry LLM benchmarks for initial guidance and gradually build an enterprise-specific benchmark for diverse set of scenarios. Industry benchmarks:

1️⃣ Based on the LLM capabilities and Enterprise alignment, build a decision matrix to help drive LLM selection within the enterprise (Reference research: arxiv.org/html/2402.06...)
www.swebench.com/

AI Bits #5 - With a wide range of LLMs available, the most common architectural decision when building Agentic AI Apps is to choose the appropriate model 👇

3️⃣ Reasoning engine (aka the "Brain "): receives feedback from the environment, self-controls and adapts its actions

4️⃣ Actuators: Action results can go back into the model, agent interacts with environment with actions too

Reference: aima.cs.berkeley.edu

AI Bits - Anatomy of an AI Agent 👇
1️⃣Sensors & Perception Layer: Information about the environment, agent interacts with environment with sensors
2️⃣ Knowledge Base (aka the “Memory”) and Feedback Mechanism