I build the data & AI layer of a MarTech SaaS — a semantic engine that powers analytics dashboards, multi-agent LLM systems that automate business workflows, and pipelines that turn raw ad-platform data into decisions.

_{Previously: co-founded product as CTO, doubled campaign ROI as a marketing data analyst. Computer Engineering @ UNIVESP.}

Server-side tracking platform — rust axum postgresql

Production SaaS ingesting, enriching and forwarding conversion events to Meta CAPI, Google Ads Enhanced Conversions and GA4 — multi-tenant, feature-flagged, built to replace client-side tracking entirely.

First-party tag SDK — javascript browser

A lightweight in-house replacement for Google Tag Manager: instrumentation rules and data extractors fetched from the platform API, deployed as a single standalone script on client sites.

Semantic analytics engine — typescript duckdb

A query-compilation layer between the warehouse and React dashboards: every metric, filter and aggregation flows through one governed engine — no ad-hoc SQL in the frontend, ever.

Multi-agent AI operations system — llm orchestration postgresql

LLM agents with persistent memory over a dedicated database cluster — they read the company's codebases, explain systems on demand and automate business workflows end-to-end.

Multi-tenant marketing data pipeline — python fastapi

Collectors for Meta, Google Ads, GA4 and RD Station APIs feeding Postgres and automated reporting — the data backbone behind campaign decisions for every client.

Operiva — rust ddd

An independent SaaS product in the making, built on hexagonal architecture: isolated domain/application/api crates, typed value objects, sqlx migrations, CI from day one.

_{These systems run in production or active development in private codebases — happy to walk through the architecture.}

LANGUAGES & CORE

python sql rust typescript postgresql duckdb fastapi react

DATA PLATFORM

airflow dbt bigquery spark kafka terraform databricks

MARTECH

meta capi google ads api ga4 sgtm server-side tracking

AI SYSTEMS

llm multi-agent orchestration rag · pgvector eval pipelines fine-tuning

reach me — LinkedIn · data.nicksonindiani@gmail.com