Leasey
Platform

The AI stack we use — and why.

We work across every layer of the modern AI stack. We pick tools to fit your constraints, not ours. Here's what we actually ship with.

Category

Model Providers

OpenAI

GPT-4.1, GPT-4o, o-series reasoning, embeddings, realtime.

Anthropic

Claude Opus, Sonnet, Haiku for reasoning, coding, and tool use.

AWS Bedrock

Managed access to Claude, Llama, Titan, and custom models in VPC.

Azure OpenAI

Enterprise OpenAI with private networking and data residency.

Google Vertex AI

Gemini, Claude, and Model Garden with GCP-native IAM.

Cohere

Command, Embed, and Rerank for retrieval-heavy workloads.

Voyage AI

Domain-tuned embeddings and rerankers.

Category

Orchestration

LangChain

Composable chains, tool wiring, and provider-agnostic glue.

LangGraph

Stateful, cyclic graph orchestration for agents and workflows.

LlamaIndex

Data framework for ingestion, indexing, and retrieval.

Temporal

Durable, long-running workflows for agent execution.

AWS Step Functions

Serverless orchestration for AI pipelines and agents.

Azure Durable Functions

Stateful serverless workflows on Azure.

Category

Vector & Retrieval

Pinecone

Managed vector database with hybrid search.

Weaviate

Open-source vector DB with modular retrievers.

pgvector

Vector search inside Postgres you already run.

OpenSearch

Hybrid BM25 + vector search on AWS.

Qdrant

High-performance vector DB with filtering.

Category

Agent Frameworks

Claude Tool Use

Native tool calling on Anthropic models.

OpenAI Functions

Structured function and tool calling.

CrewAI

Multi-agent collaboration patterns.

AutoGen

Conversational multi-agent framework.

Category

MCP

MCP TypeScript SDK

Build MCP servers and clients in Node.

MCP Python SDK

Build MCP servers and clients in Python.

Claude Desktop

Reference MCP client for local and remote servers.

Cursor

Developer IDE with first-class MCP server support.

Category

Evaluation & Observability

Langfuse

Open-source LLM observability, evals, and prompt management.

LangSmith

Tracing, evaluation, and monitoring from LangChain.

MLflow

Experiment tracking, model registry, and GenAI evaluation.

Arize Phoenix

Open-source tracing and eval for LLM apps.

Weights & Biases

Experiment tracking and eval dashboards.

Category

Cloud Platforms

AWS

Bedrock, SageMaker, Lambda, ECS, Step Functions, OpenSearch.

Azure

Azure OpenAI, AI Foundry, AKS, Functions, Cosmos, AI Search.

GCP

Vertex AI, Cloud Run, GKE, BigQuery, Model Garden.

Cloudflare

Workers AI, Vectorize, and edge inference.

Category

Data & Ingestion

Unstructured.io

Document parsing across PDF, HTML, email, and more.

LlamaParse

Layout-aware parsing tuned for RAG.

AWS Textract

OCR and form extraction for scanned documents.

Azure Document Intelligence

Layout, form, and table extraction at scale.

dbt

Analytics engineering for the warehouse behind your AI.

How we pick.

Fit the cloud you're on

Bedrock on AWS, Azure OpenAI on Azure, Vertex on GCP. Private networking and IAM you already trust.

Fit the team

If your team runs Postgres, pgvector beats standing up a new DB. We default to what you already operate.

Fit the eval

Final call goes to the numbers. We A/B models, rerankers, and chunking on your data, not vendor demos.

Ready to accelerate your tech growth?

Schedule your free consultation today and let's discuss how we can help your business scale efficiently.

Tech growth illustration
Ready when you are

Let’s ship your AI system.

Whether you’re scoping a new LLM product, hardening an existing one, or standing up the infra behind it — we’ll map the shortest path to production.

Email the teamOther ways to reach us