Leasey
Engagements

Scope we can start on Monday.

Fixed-scope, fixed-fee engagements from a 72-hour prototype sprint to a 6-week production RAG build. Real deliverables, real evals, no surprise invoices.

72 hoursPopular

72-Hour AI Prototype Sprint

From idea to working prototype in a weekend

$5,000
Fixed price, no scope creep

Validate an LLM feature, RAG use case, or agent workflow with a production-shaped prototype built against your real data. You leave with a working system, an eval baseline, and a clear path from prototype to production.

What you get

  • Working prototype against your real data
  • Eval baseline on a gold set you can own
  • Architecture doc and cost model
  • Runbook + handoff walkthrough

Ideal for

  • — De-risking an AI bet before funding a full build
  • — Getting an internal approval meeting a demo
  • — Killing bad ideas fast and cheap
72 hrs
end-to-end turnaround
Real data
not a toy dataset
Evals
baseline on day one
Book the sprint
2 weeksPopular

LLM Evaluation Audit

Prove whether your AI workflow is actually working

$25,000
Fixed scope, fixed fee

Short, intense audit of an existing LLM workflow. We build an eval harness on your real traffic, score retrieval quality, faithfulness, cost, latency, and safety, and ship a ranked fix list.

What you get

  • Versioned eval dataset built from your real traffic
  • Full eval report across 6 areas with failure-mode examples
  • Ranked fix list: biggest-impact first, with effort estimates
  • CI eval harness wired into your repo
  • Live walkthrough + Q&A with your team

Ideal for

  • — LLM systems already in production but hard to trust
  • — Pre-GA hardening of a customer-facing AI feature
  • — Regulated teams that need evidence, not opinions
6 areas
retrieval, faithfulness, safety, cost, latency, drift
Ranked
fixes, not a generic report
CI-gated
regression protection after
Start an audit
6 weeksPopular

Production RAG Build

Ship a grounded, observable RAG system in 6 weeks

From $75,000
Scoped after discovery call

End-to-end RAG build on your data and cloud. Ingestion, hybrid retrieval, reranking, generation, guardrails, observability, and a CI eval harness — shipped to production with a runbook.

What you get

  • Ingestion pipelines for your source systems
  • Hybrid retrieval with reranking and permission filtering
  • Prompt + model routing with cost controls
  • Guardrails: PII, refusal, faithfulness, jailbreak
  • Observability: traces, metrics, token + dollar attribution
  • CI eval harness and production runbook

Ideal for

  • — Teams that need a real RAG system, not a demo
  • — Regulated rollouts (HIPAA, SOC 2, GDPR)
  • — Internal copilots and customer-facing knowledge assistants
Production
grade from day one
Cloud-native
AWS / Azure / GCP
Handoff
your team owns what we ship
Scope a build
4 weeks

AI Infrastructure Foundations

Secure, observable, cost-governed AI cloud in 4 weeks

From $45,000
Scoped after discovery call

We stand up the platform layer before the first workload ships. VPC, private model endpoints, secrets, observability, cost controls, and CI/CD for prompts and models — IaC-first.

What you get

  • VPC + private networking for AI workloads
  • Private model endpoints (Bedrock, Azure OpenAI, or Vertex)
  • KMS, secrets, and key-rotation automation
  • Observability stack wired end to end
  • Budgets, quotas, and per-tenant cost attribution
  • CI/CD templates for prompts, evals, and models

Ideal for

  • — Teams kicking off their first serious AI initiative
  • — Regulated enterprises needing HIPAA / SOC 2 scaffolding
  • — Platform teams standardizing AI infra for many product teams
Day 1
compliance-ready foundations
100%
of calls traced and attributed
IaC
Terraform, CDK, Bicep, or Pulumi
Plan foundations
3 weeks

MCP Server Build

Expose your systems to any MCP-aware agent

From $30,000
Scoped after discovery call

Custom Model Context Protocol server that safely exposes your internal APIs, databases, and workflows to Claude Desktop, Cursor, and any MCP-aware agent — with auth, scoping, and audit.

What you get

  • MCP server (TypeScript or Python) for your systems
  • Auth gateway with scoped tokens + audit trail
  • Rate limits, quotas, and per-user scoping
  • Deployment to AWS, Azure, GCP, or on-prem
  • Client setup docs for Claude Desktop, Cursor, custom agents

Ideal for

  • — Teams whose users live in Claude Desktop or Cursor
  • — Platforms that want partner-facing MCP integrations
  • — Internal dev-tooling MCP servers (code, logs, runbooks)
Any client
MCP-aware agent works day one
Hours
to add a new tool safely
Full audit
of every tool call
Build an MCP server
Monthly

Ongoing AI Engineering Partnership

Embedded AI engineering, month to month

From $20,000 / mo
Retainer, cancel with 30 days notice

Act as your embedded AI engineering team — shipping features, running evals, tuning infra, and keeping the AI systems we built on a steady cadence.

What you get

  • Dedicated weekly engineering capacity
  • Continuous eval + regression monitoring
  • New feature design, build, and rollout
  • Model upgrade and cost-optimization cycles
  • Monthly written report + live review

Ideal for

  • — Teams with AI in production but no in-house AI specialist yet
  • — Founders who need a fractional AI CTO
  • — Platforms running a steady stream of AI feature work
Predictable
monthly AI engineering capacity
Evidence
every change is measured
Exit-ready
your team can take over anytime
Explore retainer

Not sure which fits?

Book a 30-minute discovery call. We'll scope the right engagement for your problem, or tell you honestly if you don't need us yet.

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