I turn messy, ambiguous problems into things that actually ship.

I'm Marco. Most of my career the title's been PM — project, program, portfolio — across software, infrastructure, M&A, and process work.

What my resume doesn't show: I've always had my hands in the work too —
business analysis, UX, some dev, testing, the data side.

I'm the one who makes sure all of it fits together and moves toward the same goal. It's one thing to have the specialists in the room. It's another to get them building the same thing. That's what I've done my whole career.

The upside: I've had a front-row seat to all of it. And with AI where it is now, I can offer more — I can help with some of the build, and help things move faster.

Instead of talking about it, here's what I mean.

See what I've built →
WHO I AM

What I'm good at is getting a room full of specialists pulling in the same direction and actually shipping. Project, program and portfolio management, agile coaching, organizing people — that's the lane I'm deep in. I've done it across public and private sector, startups to large organizations, so I know how the serious parts work — security reviews, code reviews, governance. And I've spent years on the product side, working closely with people through design thinking, which is where I thrive.

And before any of that — finding the right problem. Actually understanding what needs to be done, not just taking the request at face value. Plenty of people will gather the requirements, write the spec, and call it done. I'd rather make sure we're solving the thing that actually matters first.

I'm not the deep specialist — not the data scientist, the AI or software engineer, or the solutions architect — and I don't need to be. They go deep and own their piece; when it's done, it's done. I'm the one carrying the whole thing — the handoffs between pieces, the dependencies, the lead times — making it all fit and actually ship. I've had a front-row seat to every part of it and always had my hands in the work, so I can take a problem end-to-end — and with AI where it is now, I build more of it myself.

Work

Personal projects — built to learn, and to show what I can do.

These aren't products I'm selling. I built them to learn and to apply what I know — full, end-to-end builds from the ground up with AI: database, pipelines, backends, APIs, GenAI and LLMs, ML models, MLOps. I kept real use cases in mind the whole way, so they're close to the real thing. If you look at one and think "that's basically my problem" — that's the point. Tell me, and maybe we build the real version together.

01 / 03

StatusWatch

— a signal tool for retail investors

Background

I wanted to build something that pulls in lots of different data sources and finds the connections between them. I landed on investing — the data's easy to get, and I had a personal stake: I'd tried this myself, gone in blind, and lost money. I wanted to actually see how the day's news connects to a stock's moves, in a way I could act on.

The problem

Retail investors drown in market news and can't tell signal from noise. Most tools hand you a number and hide how they got there.

What I built

A tool that reads the day's market news, finds the connections that matter, and turns them into a clear, explained call on each stock — showing its work instead of asking you to trust a black box.

Under the hood

news sentimententity extractionentity masking for privacyembeddingsknowledge graphcorrelation engineML classifiersSHAP explainabilitymulti-horizon predictionsbacktest simulatorLLM-powered Q&A (Ask SW)multi-source ingestion pipelinemicroservicesprompt chaining & orchestrationmodel versioning & experiment tracking (MLOps)

Notable features

  • Plain-language BUY / WATCH / HOLD call per stock, with the reasoning shown — so you're never trusting a number you can't question.
  • A "gut-check" deep dive: track record plus the drivers behind each call — so you can see whether the tool's earned your trust on that stock.
  • A simulator to test a strategy against history before you risk real money.
  • Paper trading to practice acting on the signal — no real stakes.
  • Ask questions about the data in plain English.

What it proves

Taking an ambiguous problem all the way to a working product — the pipeline, the models, the explanations, the interface, all mine. If a problem like this is yours too, that's exactly the kind of thing we could build together.

Message me on LinkedIn →
statuswatch — today's predictions
12 StatusWatch home — a watchlist of tickers with plain-language verdicts and multi-horizon predictions
  1. 1A plain-language verdict per ticker — BUY / WATCH / HOLD — not just a probability.
  2. 2Conviction and other horizons (1d / 30d) surfaced right beside the call.
statuswatch — NVDA gut-check
StatusWatch NVDA detail — multi-horizon predictions, track record, and SHAP feature drivers single-ticker deep dive ↓

Per-ticker gut-check: predictions, track record, and the SHAP drivers behind each call.

02 / 03

EA Assistant

— an assistant for special-education aides
ea assistant — aide dashboard
12 EA Assistant dashboard — each student's day, schedule, and IEP goals at a glance
  1. 1Caseload, today's incidents, and open IEP goals at a glance.
  2. 2Each student's daily behavior tied straight back to the IEP goals it supports.
ea assistant — patterns
EA Assistant patterns — when, where, and how often incidents happen across a caseload

Pattern analytics across a caseload — time of day, location, severity, and trigger → behavior chains.

Background

This one started somewhere else. I was playing with image generation — building sequences of related pictures. A personal connection who works as an EA saw it and said they build visual schedules out of little picture-cards, and that what I was doing might help them. That turned into EA Assistant — for education assistants, who I think are badly underserved.

The problem

Education assistants supporting students with IEPs juggle dense plans and daily behavior logs with tools that were never built for them — and the data is sensitive.

What I built

An assistant that helps an aide actually understand a student's plan, and ties each day's behavior back to the goals it supports — built with privacy in mind from the first line.

Under the hood

voice-to-text loggingAI-suggested behavior tags"improve wording"IEP extractioncaseload pattern analyticsprompt engineeringprompt library stored in the database

Notable features

  • Reads a student's IEP and surfaces the goals being worked — so the plan isn't just a document nobody opens.
  • Ties daily behavior logs straight to those goals — so the day's work connects to the plan.
  • Voice logging — describe an incident out loud and the fields fill in — for aides documenting after a long day, or capturing it before they forget.
  • Visual schedule builder — builds picture-card routines and generates the images — the feature that started the whole project.
  • Pattern analytics across a caseload — when, where, and what tends to happen — so patterns surface instead of staying buried in logs. (The same approach could help other support roles that live between a plan and the day-to-day.)

What it proves

A real workflow for real people, where privacy and a usable interface were the hard part — and designing for that from day one was the point. If you work somewhere a tool like this is missing, let's talk.

Message me on LinkedIn →
03 / 03

Shared AI Cost Tracker

— observability for my own AI spend

Background

Building across several AI providers, my spend was invisible and scattered. So I built myself a way to see it.

The problem

No clean way to see what each project, model, or piece actually costs.

What I built

A tracker for every model call across all my projects — cost by project, model, and component, down to the cent.

Under the hood

a reusable instrumentation library shared across every projecttracks both code and automated workflowsmulti-providerper-call drill-downCSV export

Notable features

  • Spend to the cent, by project / model / component — so nothing hides.
  • Drill into every single call, group and roll up — to find the expensive thing fast.
  • Covers code and automated workflows, not just one — because cost leaks in both.
  • Export to CSV.

What it proves

I build the unglamorous infrastructure too — and I reuse it. One small library keeps a multi-project operation honest. If you've got messy AI spend you can't see, that's solvable.

Message me on LinkedIn →
ai costs — overview
12 AI Cost Tracker overview — total spend, trend, and breakdown by provider and source
  1. 1Spend tracked to the cent, with calls and average cost-per-call.
  2. 2Breakdowns by provider, source, and component — every dollar accounted for.
ai costs — explorer
AI Cost Tracker explorer — every API call, filterable by project, component, provider, and model

Explorer: drill into every call, group to roll up, and export as CSV.

Still building all of these — full backlog. If something here interests you, check back as I ship new stuff.

What I offer

Breadth is the offer.

My deep lane is leading and organizing the work — project / program / portfolio management, agile coaching, getting specialists pulling the same way. Around that I do business analysis, UX and some front-end, product and design thinking, real analysis — and now, with AI, I build more of it myself.

Want the full history? Here's my LinkedIn →