Trend Intelligence

Technology Radar

Strategic assessment of emerging and established technologies — with references to authoritative industry sources.

About This Radar

Each assessment is grounded in practitioner experience and corroborated by authoritative sources including the ThoughtWorks Technology Radar, CNCF Annual Survey, DORA State of DevOps, Gartner Hype Cycle, and Stack Overflow Developer Survey. Source links appear on each item.

Content last reviewed: September 2026

Adopt

Technologies proven in production and ready for mainstream use. Strong recommendation for most enterprises.

PostgreSQL

Open-source relational database with JSON, Full-Text Search, and pgvector. Still the #1 most admired (65.5%) and most desired (46.5%) database in 2025 — a position Stack Overflow notes it has held on both counts since 2023.

Stack Overflow Developer Survey 2025

Kubernetes

Container orchestration standard, now established as the de facto operating system for AI workloads. Production use hit 82% among cloud-native teams in 2025.

CNCF Annual Survey 2025

Terraform

Still the most widely deployed Infrastructure-as-Code tool. Note the vendor picture has shifted: HashiCorp moved Terraform off its open-source license to BUSL in 2023 and was acquired by IBM in 2025 ($6.4B) — teams wanting a vendor-neutral, community-governed option now typically evaluate OpenTofu alongside it (see Trial).

IBM–HashiCorp acquisition, TechCrunch 2025

Redis

In-memory data structure store; still the industry standard for caching, sessions, pub-sub, and real-time analytics. Took a detour to a source-available license in 2024, but reversed course in May 2025 and returned to genuine open source (AGPLv3). The Linux Foundation-backed Valkey fork, created during the interim, remains a widely used BSD-licensed alternative.

Redis relicensing to AGPLv3, 2025

OpenTelemetry

CNCF-graduated unified observability standard for metrics, logs, and traces. Replaces vendor-specific SDKs.

CNCF Landscape

Git / GitHub Actions

CI/CD via native GitHub workflows. GitHub now hosts 180M+ developers, and public projects alone burned 11.5B Actions minutes in 2025, up 35% year-over-year.

GitHub Octoverse 2025

Python

Dominant language for ML, data engineering, and backend APIs. Usage jumped 7 points year-over-year in 2025, the fastest-growing major language, driven by AI and data-science demand — even as TypeScript overtook it on raw GitHub contribution counts.

Stack Overflow Developer Survey 2025

Trial

Technologies showing strong promise. Recommended for teams with engineering bandwidth for learning and experimentation.

pgvector

PostgreSQL extension for vector similarity search. Enables AI-native workloads without a separate vector database.

GitHub (pgvector)

ClickHouse

Open-source columnar OLAP database. Processes billions of rows per second. Used by Cloudflare, ByteDance, eBay.

ClickHouse Benchmark

Databricks

Lakehouse platform. Named a Leader in the 2025 Gartner Magic Quadrant for Data Science and Machine Learning Platforms — its fourth consecutive year in that position.

Gartner Magic Quadrant 2025

Backstage

CNCF-incubated developer portal. Adopted by 2,500+ organisations including Netflix, Spotify, American Airlines.

CNCF Backstage Adopters

AI Coding Agents

AI-assisted coding (Copilot, Cursor, Claude Code) is now near-ubiquitous — 84% of developers use or plan to use AI tools, with 51% of professionals using them daily. Fully autonomous coding agents lag well behind: only 14% use them daily and 38% have no plans to.

Stack Overflow Developer Survey 2025

Temporal

Open-source workflow orchestration engine. Used by Stripe, Netflix, and DoorDash for durable, stateful processes.

Temporal Case Studies

Model Context Protocol (MCP)

Open standard, originated by Anthropic, for connecting AI applications to external tools, data, and workflows. Now supported across Claude, ChatGPT, VS Code, Cursor, and most major AI platforms — the emerging default for agent-to-tool integration.

Model Context Protocol Docs

OpenTofu

Linux Foundation-governed, fully open-source (MPL-2.0) fork of Terraform, created in direct response to HashiCorp's 2023 license change. Drop-in compatible, backed by Gruntwork, Spacelift, Harness, env0, and Scalr, and increasingly the default recommendation where license-driven vendor risk matters.

OpenTofu

Assess

Early-stage or specialised technologies. Monitor and sandbox-test. Not yet ready for critical production systems.

Agentic AI Platforms

Autonomous multi-agent systems (AutoGen, LangGraph, CrewAI). Gartner's 2026 Hype Cycle still places agentic AI at the 'Peak of Inflated Expectations': only 17% of organisations have deployed AI agents, though 60%+ plan to within two years — the most aggressive adoption curve of any emerging technology tracked. Fully autonomous agents remain unready for most enterprise use cases.

Gartner Hype Cycle for Agentic AI, 2026

Distributed SQL

NewSQL databases (CockroachDB, TiDB, YugabyteDB). ThoughtWorks recommends assessment for global-scale OLTP requirements.

ThoughtWorks Technology Radar

GraphRAG

Graph-augmented retrieval for knowledge-intensive AI. Microsoft Research published the foundational GraphRAG paper in April 2024; adoption has since spread into mainstream RAG tooling but production patterns are still maturing.

Microsoft Research 2024

Platform Engineering Portals

Specialised IDP tooling beyond Backstage (Port, Cortex, OpsLevel). Market continues to consolidate around a handful of leaders as platform engineering matures into a standard discipline.

ThoughtWorks Technology Radar

WebAssembly (Server-side)

WASM runtime beyond browsers for edge/cloud workloads. Growing steadily in specific niches — edge computing (Cloudflare, Fastly, and Akamai, which acquired Fermyon in 2025 specifically for its WASM stack) and embedded/IoT — but still short of broad enterprise production adoption.

State of WebAssembly, 2025–2026

Hold

Patterns and technologies to minimise or avoid. Significant drawbacks exist or safer alternatives are available.

Uncontrolled AI Agents

Autonomous agents deployed without human-in-the-loop guardrails. NIST AI RMF 1.0 explicitly requires human oversight for high-risk AI systems.

NIST AI Risk Management Framework

Tool Sprawl

Uncontrolled proliferation of overlapping tools increases cognitive load, security surface, and cost. DORA reports toolchain complexity as a top predictor of low delivery performance.

DORA State of DevOps 2023

Unvalidated RAG Systems

LLM systems with RAG pipelines deployed without hallucination evaluation. AWS and Google both publish evaluation frameworks to assess retrieval quality before production deployment.

AWS Bedrock Evaluation Docs

Lift-and-Shift Cloud Migration

Moving on-premises workloads to cloud without re-architecture. Gartner research shows lift-and-shift typically increases cloud spend by 30–40% versus re-platforming.

Gartner Cloud Cost Optimisation

Over-engineered Microservices

Microservices for small teams or simple domains. Amazon Prime Video's 2023 monolith reversion saved 90% of infrastructure costs and is the canonical case study against premature decomposition.

Amazon Prime Video Engineering Blog

Monolithic Unstructured Data Lakes

Large unstructured storage without governance or data contracts. Gartner predicts 80% of data lakes become 'data swamps' without proper governance.

Gartner Data Management

How to Read the Radar

Adopt

Production-proven. Strong recommendation for most enterprises. Low adoption risk.

Trial

Promising and maturing. Recommended for teams with bandwidth. Manage operational risk.

Assess

Worth monitoring. Sandbox and POC only. Not ready for critical production systems.

Hold

Avoid or minimise. Significant drawbacks, better alternatives exist, or anti-patterns.

© 2026 Reymentos Private Limited. Budhisamvad™ is a brand of Reymentos Private Limited. All content protected under copyright.