Consulting for software factories · 25–500 people

Your software factory grew. The way you run it hasn’t.

Turnover that takes knowledge with it, tech leads nobody trained, AI everyone uses their own way, and clients asking for the same work in fewer hours. In three weeks we map where your team’s capacity goes and tell you what to fix first.

No pitch deck. We’ll ask about your company and tell you honestly whether a diagnostic makes sense.

  • We speak engineering
  • People + process + AI
  • Spanish & English
Operating map · sample How one project moves through a typical software factory A workflow diagram: client, pre-sales, tech lead, dev team, QA, owner and billing. The team goes back and forth with the tech lead, the only one who knows; QA sends work back; knowledge walks out with every person who leaves; and changes pile up waiting for the owner’s sign-off. REWORK KNOWLEDGE WALKS OUT WAITING ON THE OWNER Clientscope change Pre-salesestimate Dev team2 new devs Tech leadsole expert QAreview Billingnot invoiced Ownerapproves all
  • Work flows
  • Comes back
  • Waits on someone
Illustration · fictional software factory

A-1Self-check · 2 minutes

Signs your software factory has hit its ceiling

Tick the ones you’ve said out loud this year. You’ll see right away whether the problem starts with people, process or technology, and you can send us the result as a starting point.

0 of 6 sound familiar

People
Process
Technology

Tick what sounds familiar, and we’ll show you whether it starts with people, process or technology.

A quick read of your answers, not a diagnosis.

A-2The real problem

It’s not a talent problem. It’s how the operation is built.

Most software factories attack their problems one lever at a time: hire, buy another tool, run a leadership course. Each one helps a little and none of them sticks, because the work breaks where people, process and technology meet.

Where they meet

Where the three meet is where capacity leaks.

A tech lead who is the only one who knows, an estimate nobody reviewed, AI used with no rules. The diagnostic finds these points and ranks them by how much capacity they free.

Tap a circle or a button to explore

B-1The Capacity Diagnostic

A three-week operations assessment that shows you what to fix first

No 90-page report. You get three things you can use at your next partners’ meeting.

Area ▾ Workflow ▾ ● Only where a team waits Pre-sales Tech lead QA Dev team Billing Owner TECH LEAD → DEV TEAM · TEAM WAITS Three devs waiting on answers from one lead. WHERE TO ACT: DEFINE · TRAIN

Illustrative sample from a fictional company

What we look at

  • Pre-sales & estimation
  • Sales-to-delivery handoff
  • Tech lead role & training
  • Career paths, feedback & retention
  • Onboarding new developers
  • Knowledge held by one person
  • AI in the team’s daily work
  • Code review & QA
  • Scope changes
  • Hourly, per-sprint or outcome pricing
  • Profit by project & client
  • Owner dependence

We work with what you already use (Jira, GitHub, GitLab, Azure DevOps, ClickUp, Slack, Teams, HubSpot, Harvest or Clockify and the spreadsheets in between) and we don’t resell any tool.

Fixed fee from USD 2,000 to 5,000 depending on team size. If we continue into implementation, the diagnostic fee is credited.

B-2How it works

Three weeks, without slowing delivery

Light on your leaders’ time, invisible to your clients, and booked around each team’s sprints and deliveries.

  1. Week 0

    A 20-minute call

    You tell us what’s slowing the company down. We tell you honestly whether a diagnostic is the right next step, or whether something simpler will do.

  2. Weeks 1–3

    Three weeks with your team

    Short interviews with owners, managers, tech leads, developers, QA and pre-sales, kept anonymous and never reported for groups smaller than three. We follow a few real projects end to end, from proposal to invoice, without looking at code or client data.

  3. End of week 3

    A decision, not a report

    We walk the partners through the map, the memo and the 90-day plan. You decide what to define, who to train and what to implement, with us or on your own.

C-1Why ID for Business

Built for how a software factory actually works

We speak engineering

Twenty years in software: developer, tech lead, project lead, pre-sales and professional services. We know what a sprint, a pull request and a fixed-price proposal really cost.

People, process and technology together

Not just a leadership course, not just a tool. We look at all three at once, because that’s where the operation breaks.

We build what we recommend

Most advisors stop at the slide. We can also set up the AI rules, the profitability dashboards and the integrations between your tools.

Confidential by design

Interviews are anonymous and never reported for groups smaller than three. We don’t look at your code or your clients’ data.

C-2Who you’ll work with

An engineer who has lived every stage of the software lifecycle

“Most problems in a software factory aren’t about code quality or a lack of talent: they are operational design flaws.”

I’ve spent 20 years in software experiencing that friction from every angle: shipping production code, leading engineering teams, driving critical initiatives, and closing complex integrations for enterprise accounts across the US and Latin America[cite: 3, 5, 6, 7].

I know firsthand what it costs when the best dev is promoted to lead without management training, when critical knowledge lives in only one head, when pre-sales commits to blind estimates, and when AI is adopted without quality guardrails[cite: 4, 5].

I also founded ID for Ideas, a tech community of over 80,000 professionals across Latin America and Spain where we analyze daily how teams actually collaborate and what drives true developer autonomy[cite: 1, 2].

I don’t bring generic playbooks or slide decks. I work shoulder-to-shoulder with owners and COOs on the one part of the business that rarely gets dedicated focus: how work actually moves from initial lead to final invoice[cite: 3, 4, 5].

C-3Fit

Is this right for your company?

It’s for you if

  • You run a software factory, a development consultancy or a staff augmentation company of 25 to 500 people.
  • Turnover or middle management is holding you back.
  • You want AI in your teams for real, not just paid licenses.
  • Everything ends up going through you or your partners.
  • You’d rather fix how the operation works than only hire more people.

It’s not for you if

  • You want us to find you clients or developers.
  • You want a generic leadership course.
  • You want a long report that will sit in a drawer.
  • You only need someone to implement one specific tool.

D-1What the data says

What’s happening to software factories

Less turnover than in 2021, but more pressure everywhere: pricing, AI and leadership.

What does turnover cost you?

Move the numbers to match your factory. We use Gallup’s range: replacing a person costs between half and twice their annual salary.

Leave each year

12 people

Cost to replace them

USD 216,000 – 864,000

Before counting late projects and clients who notice.

Let’s review your number in 20 minutes

Assumptions: annual salary = monthly salary × 12; replacement cost between 0.5 and 2 times annual salary (Gallup). The starting 20% turnover is the 2024 rate for Argentine software companies (OPSSI · CESSI). An estimate, not a diagnosis.

1 in 5

of the people at Argentine software companies turned over in 2024. And replacing each one costs between half and twice their annual salary.

OPSSI · CESSI, 2024 turnover · via ITSitio · replacement cost: Gallup

82%

of managers got the job with no formal management training. And 1 in 3 workers has left a job because of a bad boss.

Chartered Management Institute, 2023 · 4,500 people in the UK · via Fortune

29%

of developers trust what AI produces (down from 40%), even though 80% already use it.

Stack Overflow 2025 Developer Survey · 49,000+ responses · source

−7.1%

drop in a year in software outsourcing rates in Latin America.

Accelerance 2026 rates guide · via FullStack

52%

of people who quit say their manager or company could have done something to keep them.

Gallup

“Nobody wants to hire man-hours anymore; they want to hire solutions and products.”
Aníbal Gonda, VP of Talent at CUTI (Uruguayan IT Chamber) · Ámbito, April 2026 · translated from Spanish

Replacing a person takes months. Fixing how the work is organized takes weeks.

D-2FAQ

Questions owners ask before the first call

What kinds of companies do you work with?

Software factories, development consultancies and staff augmentation or nearshore companies of 25 to 500 people, anywhere in Latin America.

Will this slow down delivery to our clients?

No. Interviews are short and booked around each team’s sprints and deliveries. Your clients won’t notice.

Do you need to see our code or client data?

No. We look at how work moves, not at your code or your clients’ information. When we need to understand a tool, we look at it together with your team.

Can you help us bring AI into our teams?

Yes, if the diagnostic shows it’s a priority. We define where to use it, how to review what it produces, how to measure its impact and how to explain it to your clients. We’re vendor-neutral and don’t sell licenses.

Do you train tech leads?

After the diagnostic, if leadership is the priority, we build a hands-on program for your leads around the real problems in your operation, not a generic course.

Will you find us clients or developers?

No. We focus on how your operation works on the inside: people, process and technology.

How long does it take, and how much of our time?

About three weeks from kickoff to the final walkthrough. Owners and managers join a kickoff, one short interview each and the final session. Tech leads, developers, QA and pre-sales join short interviews.

How much does it cost?

A fixed fee from USD 2,000 to 5,000 depending on team size, quoted before we start. If we continue into implementation, the diagnostic fee is credited.

Is the work remote?

Yes. Interviews and sessions run over Zoom, Google Meet, Microsoft Teams or Slack huddles, in Spanish or English. We’re based in Montevideo, Uruguay (UTC−3).

What happens after the diagnostic?

You keep the map, the memo and the 90-day plan. You can run it with your own team or ask us to help train leads, define processes, set AI rules or build dashboards. No obligation either way.

Next step

Start with a 20-minute conversation

Tell us what’s holding your software factory back. If a diagnostic isn’t the right next step, we’ll say so and point you to something simpler.

  • 20 minutes, on video
  • No preparation needed
  • A straight answer on fit

Or send a WhatsApp message

Your role
Company size
Open in WhatsApp

Opens WhatsApp with this text. Nothing is sent until you press send there.

Book a 20-min call WhatsApp