Samuel Eguino · Integration engineering for B2B SaaS

the_problem

The deal stalls in procurement. The ticket rots in the backlog.

Your enterprise prospect loves the product. Then their IT team asks whether it syncs with the HRIS they bought in 2011, and "when do we start" becomes "we'll circle back." Your engineers could build it, but every sprint they choose the roadmap over the plumbing. Reasonable choice. The deal is still stalled.

Run the math on that: somewhere in your pipeline, a six-figure contract is waiting on a five-figure integration. Every quarter it sits there, the discount to close it grows.

That plumbing is my whole practice, not my side quest. Ingestion from hostile systems, entity matching across dirty data, sync that survives schema drift and the customer who renames every column. Exciting integrations page people at 2am. Mine are boring, on purpose.

rates

Three ways to hire me. Real prices on all of them.

[1]

Unified API implementation

Finch, Merge, or Paragon wired into your product properly: auth flows, webhooks, data model mapping, failure handling. The fast path when a platform covers your target systems.

$8–12k3–4 weeks · fixed

[2]

Custom & long-tail integrations

Everything the platforms don't cover: vertical systems, on-prem exports, SFTP feeds, file-based ingestion, and the reconciliation that turns five messy sources into one clean contract.

$10–20k4–6 weeks · fixed

[3]

Maintenance retainer

Integrations rot: source systems change their exports, new customers arrive with new formats. I monitor, patch, and extend what I build so it stays boring. Priority response included.

from $2,500/moongoing

start_here

Every engagement starts with a one week paid audit ($2,500). I map your systems, data, and failure modes, then deliver a build spec with a fixed quote. Hire me for the build and the audit fee comes off the price. Or keep the spec and build in-house: it's written so any strong engineer can execute it.

Want proof before paying anyone anything? Email me and I'll send a sample audit spec so you can judge the work before the first call.

process

Audit, build, run. No hourly billing.

week 1

Audit

I map the systems, the data, and everything that can break, then hand you a spec with a fixed quote.

weeks 2–6

Build

Fixed price, fixed scope, in your repo from day one. Weekly demos of real data actually moving.

ongoing

Run

Handover with docs, monitoring, and runbooks. Or a retainer, if you'd rather never think about it again.

objections

The questions you're already asking.

Why not just use Merge or Finch directly?

Often you should, and I'll tell you when. I implement the platforms when they cover your systems (service №1) and hand-build when they don't. The audit's job is to tell you which situation you're in before you spend real money.

Can't our own engineers build this?

Yes, given time, and the audit spec is written so they can. The question is whether the integration ships this quarter or keeps losing sprint planning to the roadmap. You're not buying ability. You're buying certainty and a date.

You're one person. What if you disappear?

Everything happens in your repo from day one: code, docs, runbooks, monitoring. If I got hit by a bus in week three, your team keeps the spec, the working branches, and the paper trail. No lock-in is the design goal, not a concession.

How do you handle our customers' data?

Least privilege, your infrastructure, your credentials vault, NDA before the audit. I work in healthcare-adjacent data at my day job; paranoia is the baseline, not an add-on.

What does "fixed price" mean when scope changes?

The audit defines scope in writing before I quote. If we discover something mid-build that genuinely changes the shape of the work, you get a written change order with a price before any new work starts. No surprise invoices, ever.

about

Hi, I'm Sam. This is literally my day job.

I'm a software engineer at a healthcare SaaS, where my job is exactly this: building ingestion and intelligence layers on top of UKG, Symplr, and customer CSV exports that were never meant to leave Excel. I've seen what breaks at 2am, so I design for it at 2pm.

I take a small number of projects at a time and say no to work I can't do well. If your problem is hard and interesting, it moves up the queue.

$ ls ./systems/ ukg_proworkdaynetsuitesymplradpsftp_feedsfinchmergeparagonroster_v7_FINAL(2).csv

sam integration engineer · est. 2026

one_more_thing

Bring me the ugliest system you have.

Twenty minutes. I'll tell you honestly whether it's a quick job, a real project, or a bad idea, and what I'd do first either way.

[ book intro call ]

prefer email? samueleguino97@gmail.com