Experience
I started coding because it seemed like the one door nobody could close on me, and I’ve spent the years since showing up to things a little early. Voice AI before the LLMs. Remote work as a job filter right as it went permanent. RAG before it had an acronym worth dropping. I’m not a fortune teller. I just like building the unglamorous plumbing that turns a good idea into something people can actually use, and I’ll wander up and down the stack to do it. Here’s the work, newest first.
DoorDash
Working on the systems that convert DoorDash users into DashPass subscribers. Checkout is the highest-intent surface in the product, so my team decides what offer to show, when, and how to make it feel like a natural part of the flow. I also support subscription campaigns across Canada, Australia, and New Zealand, which means building for multi-currency, localized promotions, and different regulatory environments from a single platform. I mentor newer engineers and shape direction through RFCs and operational readiness reviews, because the systems outlive whoever shipped them.
A year before the subscription work, I was further up the funnel, making sure people could find DoorDash at all.
Built a backend-driven system for rendering internal links across doordash.com, replacing a static setup where footer links and SEO pages were maintained by hand. ETL jobs compute the links offline, things like the most popular cuisines in a submarket or the top businesses by order volume, and the framework now runs on the homepage, store and business pages, and nearly all logged-out surfaces. I also built neighborhood pages for metros like New York, LA, and Chicago, dedicated pages for places like Chelsea or Silver Lake featuring local restaurants. Along the way I fixed price inconsistencies between the cart and checkout pages that were confusing customers, and ported high-performing iOS and Android features to web.
Parente AI
Between LinkedIn and DoorDash I consulted on a psychology-focused chat application. Built a CMS, integrated a vector database with embeddings, and implemented retrieval-augmented generation in the chat interface, right before RAG became the industry’s favorite acronym.
What is RAG? - AWS
Retrieval-augmented generation explained, the architecture pattern behind modern AI applications
aws.amazon.comBefore the consulting work, I spent three years at LinkedIn, where I learned what timing does to a product.
Joined right as remote work stopped being a temporary arrangement and started being a permanent feature of the labor market. My team added Remote and Hybrid as first-class options in LinkedIn Recruiter’s ATS. Remote roles became filterable for the first time, so applicants could actually find them. Built the admin panel for the Job Wrapping product, which integrates LinkedIn job listings with third-party ATS platforms like Greenhouse, Ashby, and Lever. Improved the Job Management experience so recruiters could upgrade or downgrade postings between paid and free without friction across millions of active listings.
On the infrastructure side, I migrated the localization system to Ember 4.0, deprecated legacy APIs, and transitioned surfaces from Rest.li to GraphQL. The kind of modernization work that unblocks everyone else. Led technical direction through RFCs, worked with Accessibility and Compliance on audit findings, mentored two contractors, and conducted candidate interviews.
Before recruiting, the users were smaller and the rules were bigger. I was building for kids.
Meta
Led the migration of messengerkids.com from an external vendor to Facebook’s internal CMS, built with React and Hack. What had been a slow, vendor-dependent update cycle moved in-house, so changes shipped on our own schedule. Integrated the CDN for streaming video, documents, and images. Added localization and A/B testing support so the site could serve more regions and experiment with content safely.
The part that shaped me most was the compliance work. GDPR Article 8 sets consent thresholds for children that vary by EU member state, so localization wasn’t a nice-to-have, it was a legal requirement. Years before children’s safety online became front-page policy, working on COPPA and GDPR compliance for a kids’ product taught me that privacy is an engineering requirement, not a checkbox.
Before the recognizable logos, there were the small rooms where I learned to ship, usually a little early.
Startups & Consulting
Built voice AI that automated restaurant phone orders in 2019, years before LLMs made “AI taking your order” a headline. NLU pipelines, intent classification, and a lot of domain-specific tuning. Restaurant voice AI is a crowded category now. In 2019 it wasn’t, and we were already building it.
Cybersecurity startup doing edge AI for threat detection. Built internal tools and dashboards, integrated the MITRE ATT&CK framework, and set up physical server racks. MistNet was later acquired by LogRhythm.
Apprentice at an early-stage proptech startup in private real estate investing. First professional engineering role. Learned to write production code while the company searched for product-market fit.
Recommendations
What colleagues have said, pulled from LinkedIn. Two themes keep repeating: works across the whole stack, and leaves people better than he found them.
It was a pleasure working with Carlos at DoorDash. He was one of the few engineers who worked across the stack, and created and owned multiple pages from the Database layer all the way up to React components. Aside from working with PMs and Analytics, Carlos also made time to set up team games, knowledge sharing sessions, and offsite activities with engineers, making sure we all felt included and had a good time. I am sure of him being a valued asset on any team lucky to have him.
I am excited to share my experience being mentored by Carlos during my software engineering internship at DoorDash. He provided me an incredible amount of support throughout the summer; not only did he help me build my technical knowledge, but his thoroughness, approachability, and attention to detail gave me an understanding of what makes a successful engineer and team member. Carlos cares deeply about the work he does, and his choice to become a mentor on top of his regular responsibilities is a reflection of this. Any team would be lucky to have him!
Carlos is a great learning-focused engineer who always pushes the boundaries of understanding how to write the best code possible. I would gladly work with him again in the future!
I am writing to wholeheartedly recommend Carlos for any engineering position. I have had the pleasure of working with Carlos at LinkedIn, and I can confidently say that he is one of the most talented and dedicated engineers I have ever met. Carlos possesses a deep understanding of technical concepts and demonstrates exceptional problem-solving skills. He has consistently resolved critical issues with remarkable speed and efficiency, often under tight deadlines. Carlos is an invaluable asset to any team, not only for his technical prowess but also for his positive attitude and exceptional communication skills.
I'm lucky to have met such an amazing colleague who's so supportive and a good listener. Carlos makes you want to be a better version of yourself and challenges you to think in new and creative ways. I'm always in awe and inspired by the way he operates as well as his technical depth in all things frontend, Javascript, and software engineering. He has this amazing curiosity which goes beyond getting to the root cause, which is so rare to see in an engineer. On top of all this, he's an excellent mentor and always provides knowledge, guidance, and resources on the drop of a penny.
Carlos is a good UI engineer and a great team player. His out-of-box thinking and knack for learning and applying new technologies to his work is commendable. He worked on several key features related to Job Wrapping 3.0 like Admin rule management dashboard, Self serve onboarding, and JW banner. Carlos has good leadership skills and has helped a couple of contractors come up to speed quickly on projects related to LinkedIn Jobs products.
Carlos and I worked on the same team at LinkedIn as front-end engineers and in addition to that he also was my mentor, and the kind of mentor one can only wish for. I admire his javascript knowledge and am very thankful for the opportunity to work with such an amazing engineer. Without a doubt, Carlos will be an amazing addition to any team he'll join.
Carlos was incredibly hard working, bright and gave honest credible advice on concepts he understood. He picked up on new industry skills incredibly quickly and was extremely effective with Micro-service Development, Helping other Software Engineers and contributing to the team as a whole. His technical skills and amazing communication skills were a real asset to us at the early stages of Rettex.
Education
A project-based CS program backed by Y Combinator, built around shipping real products instead of passing theory exams. I didn’t finish, though. By then I’d already fallen down the rabbit hole of building things on my own, and the rest I taught myself. Make School, a startup teaching at a startup, has since been absorbed into Dominican University of California.
I learned engineering the way I learned to cook and fix bikes at home: because the door was open and nobody was going to close it for me. That part hasn’t changed. Still early, still figuring it out, still building.