⚙️
Enterprise & startups
Comfortable in regulated, large-scale codebases and in zero-to-one startup builds where I wear every hat.
Available for remote contract work
Jeremias Aguero
Senior Full-Stack & AI Engineer — building AI-native products, end to end.
|
7+ years shipping production web applications — from architecture and API design through deployment, monitoring, and scale. Recently focused on turning LLM features, RAG pipelines, and vector search into monitored, production-grade systems.
from langchain import hub
from langchain_postgres import PGVector
# retrieve, then ground the model
store = PGVector(embeddings, collection="docs")
docs = store.similarity_search(query, k=6)
answer = llm.invoke(
prompt.format(context=docs, question=query),
).content # streamed + traced in prod
0
Years shipping
0
Years with U.S. teams
0
Products shipped
0
End-to-end ownership
01 — About
I take products all the way.
I'm a Senior Full-Stack Engineer with 7+ years building and shipping production web applications — including 6+ years with U.S. companies across both enterprise platforms and fast-moving startups.
I work primarily in TypeScript/JavaScript, Node.js, and Python, with React and Next.js on the front end. I take products end to end — from architecture and API design through deployment, monitoring, and scale — and I can own a feature from requirements to live deployment independently.
Lately my focus has been integrating AI into real products: LLM-powered features, RAG pipelines, and vector search — taking them from prototype to monitored, production-grade systems rather than one-off API calls. I care about measurable outcomes: faster load times, lower infrastructure cost, and fewer defects.
🧠
AI in real products
LLM features, RAG, and vector search taken from prototype to production — traced, evaluated, and monitored.
🚀
End-to-end delivery
From whiteboard to live deploy: data modeling, APIs, UI, infra, observability, and the rollout in between.
🌍
Remote-ready
Open to remote contract work with global teams. Comfortable working across U.S. time zones.
02 — What I bring
A bias toward measurable outcomes.
End-to-end ownership
I take a feature from requirements to live deployment independently — data model, API, UI, infrastructure, and rollout. No hand-offs required.
Production-grade AI
RAG pipelines, vector search, and LLM features built with retrieval quality, evaluation, guardrails, and tracing — not brittle prompt hacks.
Scalable architecture
API and system design that holds up under load — sensible boundaries, caching, queues, and databases chosen for the actual workload.
Measurable impact
A bias toward outcomes you can point to: faster load times, lower infrastructure cost, and fewer defects in production.
Clear communication
Collaboration across product, design, and stakeholders — async-friendly, well-documented, and decisive so the right thing ships.
Quality & reliability
Tested, observable systems — automated testing, monitoring, and CI/CD so changes ship fast and stay stable.
03 — Core stack
Tools I reach for, production-tested.
04 — Selected work
A few things I've shipped.
🧩
GitHub ↗
Fullstack AI SaaS Starter
A TypeScript-first, type-safe SaaS boilerplate with AI built in — Fastify + tRPC backend, React front end, PostgreSQL, and OpenAI-powered streaming chat over server-sent events.
TypeScript
Fastify
tRPC
React
PostgreSQL
OpenAI
🎬
GitHub ↗
YouTube Summarizer
Auto-detects chapters and topics in YouTube videos, generates summaries in your chosen language, and adds clickable timestamps that jump straight to the moment.
Next.js 16
React 19
TypeScript
Prisma
GLM-4.7
Tailwind
🔗
GitHub ↗
LangChain Examples
A collection of practical LangChain apps — document summarization, retrieval Q&A, and web-search agents — built on OpenAI and Gemini with Pinecone and Chroma vector stores.
Python
LangChain
RAG
Streamlit
Pinecone
Chroma
🦛
GitHub ↗
Hippo Memory
A biologically-inspired memory layer for AI agents: memories decay by default and strengthen on retrieval, so agents stop repeating mistakes. Zero-dependency SQLite store exposed over MCP.
TypeScript
Node.js
SQLite
MCP
Embeddings
05 — Contact
Let's build something production-grade.
Open to remote contract work with global teams. If you're integrating AI into a real product and want it monitored and scalable from day one — let's talk.
jeremiasaguero0323@gmail.com