Fixed-scope software delivery

Get the engineering work off your backlog without opening a full-time role.

When a feature, integration, AI backend, demo, or technical decision keeps slipping, Montpelier Consulting turns it into a quoted scope and a scheduled delivery. Adam Pippert defines the boundary, prices the work, and owns the client conversation. Two bench engineers are available for implementation, so useful work can move without waiting for one calendar to clear.

Send the stuck project Pick an entry point

Backlog itemA narrow build, fix, integration, or review that needs an owner.

Written quoteScope, price, timeline, exclusions, and acceptance criteria before work starts.

Accountable leadAdam scopes and manages delivery instead of handing you anonymous labor.

Why teams call

The work is too important to keep delaying and too bounded to justify a hire.

Small teams rarely need “more engineering capacity” in the abstract. They need one thing finished: the brittle integration, the internal workflow, the AI feature that needs a real backend, the demo sales keeps promising, or the architecture decision nobody has time to settle. Open-ended contractors turn that into hourly drift. Montpelier turns it into a boundary.

Bring the business outcome

“Fix this flow,” “build this feature,” “ship this demo,” or “tell us if this architecture is safe.”
The repo, API, workflow, constraints, examples, or stakeholder context needed to draw the line.
A decision-maker who can agree to acceptance criteria before implementation begins.

Get the delivery boundary

A written scope with inclusions, exclusions, timeline, fixed price, and review criteria.
A technical lead who prices the work and remains accountable for the result.
A concrete artifact: code, fix, integration, report, diagram, prototype, or proposal-ready technical material.

Best-fit work

Use Montpelier when the next step is concrete enough to quote.

The best projects have a visible owner, a useful deadline, and a definition of done. Montpelier is not trying to become your permanent engineering department. It is built to take a bounded request, make the tradeoffs explicit, and deliver the agreed artifact.

Delivered

Software engineering & technical presales

Features, integrations, internal tools, demos, technical scoping, proposal support, and sales-engineering-adjacent deliverables. This is the current strongest sell because it is backed by a paid $10,000 engagement.

Demonstrated

AI infrastructure & RAG systems

Retrieval pipelines, citation-grounded answer flows, streaming responses, and evaluation harnesses for teams whose chatbot prototype needs to stop inventing unsupported answers.

Demonstrated

Agent orchestration backends

Request routing, tool-use coordination, retries, fallbacks, failover paths, and streaming state updates for teams moving from “the agent demo works” to “customers can rely on it.”

Selective / learning track

Legacy modernization & inference optimization

These tracks are in active development. If your project touches old systems or model-serving cost, expect a direct conversation about what is proven, what is demonstrated, and what still needs discovery.

Delivery model

You get bench capacity without becoming the project manager.

The bench matters because it gives your project implementation capacity. The accountability matters more. Adam scopes and prices the job, decides which work should be delegated, and remains your point of contact. The engineers help deliver the scope; they do not turn your request into a staffing exercise.

STEP 1

Convert the request

Your request becomes a written scope: deliverables, timeline, assumptions, exclusions, and acceptance criteria.

STEP 2

Assign the right capacity

Adam leads the work. A bench engineer takes implementation when that is the fastest safe path. Complex, sensitive, or ambiguous work stays closer to Adam.

STEP 3

Review the agreed result

You review the result against the agreed criteria. New requests become a separate priced change, not surprise hours on the same invoice.

Fixed-price entry points

Buy the smallest useful next step.

A good first engagement should reduce uncertainty, not create a retainer you have to manage. Start with a working session if you need a decision, a fix-it sprint if the task is already known, an audit if the system needs review, or a build sprint when the artifact is defined.

Technical Working Session

$250

60-90 minutes, usually same week

Use this when you need a technical read before committing: architecture gut-check, code review, debugging second opinion, build-vs-buy decision, or proposal scoping. You leave with 3-5 written follow-up bullets.

Book the $250 session

Fix-It Sprint

$400

1-2 business days

Use this for one narrow, already-understood task: a bug, script, one-endpoint integration, config problem, or deployment fix. No broad discovery. No vague backlog cleanup.

Start the $400 sprint

Package	Best when you need	Timeline	Price
Discovery & Audit Sprint	A codebase, architecture, or technical process reviewed with a prioritized report	3-5 business days	$800
Build Sprint	One defined feature, integration, prototype, or internal tool delivered end to end	2-3 weeks	$4,000
Defined-Scope Project	A larger multi-part delivery with its own scope document and acceptance criteria	4-6 weeks	$8,000
Monthly Delivery Block	Reserved recurring capacity for ongoing improvements after the first scope works	Ongoing	$1,400/month

Inside the agreed scope, the price is fixed. Outside the agreed scope, the work is re-scoped and priced before it starts. Specialized AI infrastructure, agent orchestration, legacy-system, or inference-cost work may price differently after scoping.

Proof without pretending

Inspect the thinking before you buy the work.

Montpelier has one real paid $10,000 engagement to point to today. The downloadable projects below are different: fictional engagements built with synthetic data. They are not client case studies. Their job is to show how a request becomes a scope, an architecture, failure handling, and runnable code.

AI infrastructure & RAG

Regulatory intelligence RAG platform

Hybrid retrieval, citation-grounded answers, and a confidence floor that refuses to invent a source when the corpus does not support the answer.

Download sample (.zip)

Legacy modernization bridge

RPG II policy system to REST API

A synthetic fixed-width legacy file bridged to a modern API layer, with named failure handling for malformed records and timeout paths.

Download sample (.zip)

Agent orchestration

Multi-agent reliability audit

A task router with retry, fallback, failover, and escalation paths so an agent workflow has a plan when tools or models fail.

Download sample (.zip)

Inference optimization

Task-aware model routing & cost model

A before/after comparison between naive single-model routing and task-aware routing with a budget-threshold fallback.

Download sample (.zip)

Objections worth answering

Before you send the request

What if the work runs longer than expected?

The price does not change inside the agreed scope. Montpelier absorbs the estimate risk. Work outside the scope becomes a separate change order with its own price and approval.

What if I only have a vague problem?

Start with a working session or Discovery & Audit Sprint. The goal is to turn the vague problem into a decision, report, or quote — not to pretend an unclear build is ready for a fixed price.

Why not just hire a freelancer directly?

You can. The value here is scoping, accountability, and delivery management: one technical lead turns the request into a bounded project, assigns bench capacity where it fits, and stays responsible for the result.

Why should I trust an early-stage practice?

Because the page does not hide what is proven. One real $10,000 engagement exists. The rest is labeled as demonstrated or in progress. If you need a firm with a long public client roster, this is not that pitch. If you need a scoped piece of work with clear acceptance criteria, send the request.

How fast can you start?

A working session or fix-it sprint can usually start the same week. Larger work starts after the scope is written, the start date is confirmed, and the deposit clears.