The stack behind the work.

Node.js

JavaScript runtime for high-performance APIs and real-time services — the standard backend for TypeScript teams and AI-augmented serverless functions.

Microsoft .NET

Microsoft’s enterprise platform for C# — high-performance APIs, Blazor, Azure Functions, and MAUI spanning web, cloud, and cross-platform applications.

React Native

Build high-performance iOS and Android apps from a single JavaScript codebase — with a truly native feel.

Flutter

Google’s UI toolkit for building natively compiled apps across mobile, web, and desktop from a single codebase.

Docker & Kubernetes

Containerised deployment and orchestration — consistent environments from local dev to production at scale.

Drupal

Enterprise CMS for complex digital platforms — multilingual support, granular access control, and scalable content architecture for government and media.

WordPress

Powers 43% of the web — headless REST and GraphQL APIs, WooCommerce, and an extensive plugin ecosystem for rapid content platform deployment.

Claude Agent SDK

Anthropic’s SDK for production Claude agents — tool use, MCP integration, computer use, and multi-turn orchestration built-in.

OpenAI Agents SDK

OpenAI’s framework for multi-agent systems — agent handoffs, tool execution, guardrails, and tracing for GPT-4o and o-series models.

Gemini

Google’s multimodal AI model for text, image, and code — integrated across Google Workspace and cloud services.

Higgsfield AI

AI-powered video generation with cinematic motion control — built for high-fidelity creative and marketing content.

Kling AI

State-of-the-art text-to-video and image-to-video model — known for fluid motion and photorealistic output.

Seedance

ByteDance’s AI video generation platform — fast, scalable video creation with precise temporal and motion control.

AWS Bedrock

AWS-managed access to frontier models — Anthropic, Meta, Mistral — with built-in guardrails, RAG support, and enterprise data controls.

Azure AI Foundry

Microsoft’s platform for deploying GPT-4o, o-series, and Phi models with prompt flow tooling, evaluations, and Azure-native security.

Google Vertex AI

Google Cloud’s AI platform — Gemini, Gemma, model garden, vector search, grounding, and production MLOps pipelines in one place.

LangChain

The standard LLM application framework — 500+ integrations, composable chains, and LCEL pipelines for production RAG and agent systems.

LangGraph

Stateful agent orchestration using graph-based execution — cyclic workflows, human-in-the-loop, and checkpointing for production LLM systems.

LlamaIndex

Data framework for connecting LLMs to external sources — advanced indexing, hybrid retrieval, and query engines for enterprise RAG systems.

Semantic Kernel

Microsoft’s LLM SDK for .NET, Python, and Java — plugin architecture, planner, memory connectors, and native Azure AI Foundry integration.

CrewAI

Role-based multi-agent orchestration — assign personas, goals, and tools to autonomous AI crews for parallel task execution and delegation.

LiveKit

Real-time WebRTC infrastructure for voice and video AI agents — sub-200ms latency pipelines for production voice AI and multimodal interfaces.

Pinecone

Managed vector database for AI — millisecond similarity search at billion-vector scale with serverless architecture and multi-tenant namespacing.

Qdrant

Rust-powered open-source vector DB — dense, sparse, and hybrid search with on-premise deployment for full data sovereignty and production RAG.

pgvector

PostgreSQL extension for vector similarity search — HNSW indexing and hybrid BM25 retrieval without a separate vector store infrastructure.

HuggingFace

The open-source AI hub — 500K+ models, datasets, and the Transformers library powering fine-tuning and inference across all major architectures.

Unsloth

Fine-tuning LLMs on consumer GPUs — up to 80% less memory and 2x faster training for Llama, Gemma, Qwen, and Mistral via LoRA and full fine-tune.

RunPod

Fine-tuning LLMs on consumer GPUs — up to 80% less memory and 2x faster training for Llama, Gemma, Qwen, and Mistral via LoRA and full fine-tune.

PyTorch

The dominant deep learning framework — powers fine-tuning, custom model training, and is required in 85%+ of ML engineer job descriptions.

TensorFlow

Google’s ML framework for production at scale — TF Serving, TFX pipelines, and TF Lite for edge deployment across enterprise ML workflows.

DeepEval

LLM evaluation framework for RAG and agents — contextual recall, faithfulness, hallucination detection, and CI/CD integration for regression testing.

Opik by Comet

LLM observability platform — prompt versioning, trace inspection, experiment tracking, and evaluation dashboards for production GenAI workloads.

LangSmith

LangChain’s observability layer — full trace visibility, eval dataset management, prompt playground, and regression testing for production LLM apps.

Midjourney

Benchmark AI image generation — photorealistic and artistic visuals used by studios and agencies for concept art, brand imagery, and creative direction.

TestRigor

AI-powered test automation in plain English — self-healing tests, visual testing, and cross-platform coverage with minimal programming overhead.

Testim

AI-driven test automation with smart locators and self-healing for dynamic web apps — codeless and coded authoring with deep CI/CD and Salesforce support.

Azure Synapse Analytics

Microsoft’s unified analytics platform — data warehousing, big data, and ML pipelines with native Azure Data Lake, Power BI, and Purview integration.

AWS Glue

Serverless ETL on AWS — auto-generated PySpark code, data cataloguing, schema evolution, and 70+ source connectors for managed data pipelines.

Databricks

Unified lakehouse platform — Apache Spark, Delta Lake, MLflow, and Unity Catalog for collaborative data engineering and production ML at scale.

Snowflake

Cloud-native data warehouse with separated compute and storage, zero-copy data sharing, and Cortex AI for LLM-powered analytics at enterprise scale.

Neo4j

Leading graph database for connected data — powers knowledge graphs, fraud detection, and HybridRAG architectures combining vector and graph retrieval.

ClickHouse

Column-oriented OLAP database for real-time analytics — sub-second queries at petabyte scale with vectorised execution and exceptional compression.

PostgreSQL

Advanced open-source relational database — JSONB, full-text search, partitioning, and pgvector extension for AI-ready production workloads.

SQL Server

Microsoft’s enterprise RDBMS — native Azure Synapse, Power BI, and .NET integration with ColumnStore indexes for analytical workloads at scale.

Apache Kafka

Distributed event streaming platform handling trillions of events daily — the industry standard for real-time pipelines, CDC, and event-driven architectures.

Apache Flink

Stateful stream processing with exactly-once semantics and event-time handling — the production choice for true real-time analytics beyond micro-batching.

Apache Spark (PySpark)

Unified distributed analytics engine — PySpark brings Python-native batch and streaming computation powering the majority of enterprise data platforms.

Debezium

Open-source CDC platform capturing row-level changes from Postgres, MySQL, and SQL Server — publishes to Kafka for event-driven sync without polling.

Redpanda

Kafka-compatible event streaming in C++ — 10x lower latency, no JVM, and simpler operations as a drop-in replacement for production Kafka workloads.

Apicurio Registry

Open-source schema registry for Avro, JSON Schema, and Protobuf — enforces schema compatibility and versioning across Kafka-based data pipelines.

GitHub Actions

GitHub-native CI/CD — event-driven pipelines for build, test, deploy, and AI model evaluation with 20,000+ community actions and matrix builds.

Terraform

Infrastructure as code for AWS, Azure, and GCP — declarative HCL, state management, and 3,000+ providers make it the universal IaC standard.

Prometheus

Cloud-native metrics and monitoring — pull-based collection, PromQL queries, and Kubernetes service discovery. The CNCF standard for production observability.

Grafana

Unified observability dashboards for metrics, logs, and traces — integrates with Prometheus, Loki, Tempo, and 150+ sources from startup to hyperscale.

OpenTelemetry

CNCF open standard for distributed traces, metrics, and logs — vendor-neutral instrumentation increasingly mandated in enterprise cloud architecture.

Docker

Industry-standard containerisation for reproducible environments — the foundational deployment primitive in backend, DevOps, and AI engineering stacks.

Kubernetes

De facto container orchestration for production workloads — CKA certification demand surged 68% in 2025, making it non-negotiable for cloud-native teams.

React

Component-based JavaScript UI library with 47% framework market share — the foundation for Next.js, React Native, and modern web application frontends.

Angular

Google’s TypeScript framework for enterprise SPAs — dependency injection, RxJS reactivity, and Angular Universal for teams requiring strict architectural standards.

FastAPI

High-performance async Python API framework with automatic OpenAPI docs and Pydantic validation — the default for LLM services and RAG API deployments.

GraphQL

API query language for precise data fetching — typed schema contracts, subscriptions, and data federation eliminating over-fetching in complex integrations.

Playwright

Microsoft’s E2E browser automation — cross-browser Chromium, Firefox, WebKit with parallel execution, network interception, and visual comparisons.

Cypress

JavaScript-native E2E testing with real-time execution, time-travel debugging, and automatic waiting — fast feedback for modern frontend CI pipelines.

Selenium

The cross-browser automation standard — WebDriver protocol underpins enterprise QA infrastructure and remains a baseline in QA engineer job requirements.

Postman

API design, testing, and documentation — from manual exploration to full automated collection runs.

Newman

CLI runner for Postman collections — used to automate 500+ API tests in CI pipelines with detailed HTML reports.

BrowserStack

Real-device cloud testing across 3,500+ browsers and devices — supports Selenium, Playwright, and Cypress without emulators or physical device labs.

Jira

Bug tracking, sprint management, and test cycle reporting — the backbone of every client QE engagement.

React

JavaScript library for building fast, component-driven user interfaces — the dominant choice for modern web applications.

Next.js

React meta-framework with server-side rendering and static generation — optimised performance out of the box.

Angular

Google’s opinionated TypeScript framework for building scalable, enterprise-grade single-page applications.

Blazor

Microsoft’s web framework for building interactive UIs with C# and .NET — no JavaScript required.

Python

The lingua franca of AI and data engineering — present in 89% of AI job postings in 2026 and powering the entire modern ML and LLM stack.