DATA ENGINEERING · AI · AUTOMATION · 6 PRODUCTS LIVE

Your pipelines break at 3am. I make sure they don't.

8+ years building and fixing data infrastructure, AI systems, and automation for companies that can't afford unreliable data. I also build my own products — 6 live in production right now.

See what I've built → LinkedIn
Hear my story · 1 min
0:00
8+
years in data & AI
50+
projects shipped
$2.5M+
documented savings
6
own products live

What I solve.

These problems happen every day. I've fixed each one more than once.
01 · DOCUMENTS
What if your documents loaded themselves?
OCR + AI extract and validate automatically. I built this for an insurance client — went from 8 days of manual loading to 2 hours. Zero typos.
02 · REPETITIVE INQUIRIES
What if every repetitive question answered itself?
AI assistants connected to real data. I built one for a property manager — she recovered 25 hours/week she spent answering "how much do I owe?" on WhatsApp.
03 · SCATTERED DATA
What if you could see everything in one place?
Pipelines that unify 8 systems into one reliable warehouse. I redesigned this for a logistics company — infra costs dropped 46%, Excel dependency gone.

I don't just talk about data — I ship products.

6 products I designed, built, and operate. All live in production.
● LIVE
Properties
SaaS for rental property managers. Self-service portal, receipt OCR, AI assistant. 3 live clients, 339 tests.
See product →
● LIVE
Doc Intelligence
Upload any document — invoice, claim, contract. Structured data in 10 seconds. No templates, no setup.
See product →
● LIVE
Hotel Intelligence
Tracks every charge, detects anomalies, alerts before month-end. One hotel recovered $2K/month in untracked consumption.
See product →
● LIVE
Try-On AR
Virtual eyewear try-on. Computer Vision detects your face in real time and overlays 3D frames.
See product →
● LIVE
LanchaYa
Natural language boat search. Tell it what you need and the map shows you options.
See product →
● LIVE
Vibe Check
AI-built app reviews. Security, database, deploy — 38 points audited in 5 days.
See product →

Results with real metrics.

Own products and client work. The numbers speak.
● OWN PRODUCT
25→0 hrs/wk
Repetitive inquiries eliminated
Property manager with 200 tenants spent 25 hrs/week on WhatsApp. Now the AI assistant handles everything. She touches nothing.
OCRLLMRAG
● OWN PRODUCT
15 min → 10 sec
Document data extraction
Copying invoice data by hand took 15 minutes per document. Now upload the PDF and get everything structured in 10 seconds.
Claude APIOCR
● OWN PRODUCT
$2K/mo
Unrecorded consumption recovered
Hotel was losing ~$2,000/month in untracked consumption. Now every item is recorded and anomalies are detected instantly.
LLMQRAnomaly
INSURANCE · CLAIMS
−65%
Claims resolution time
Mid-market insurer processed claims in 8 days. Now under 3. Same team, triple capacity.
OCRNLP
HEALTHCARE · BILLING
15% → 3%
Medical billing errors
Hospital was losing thousands in rejections due to billing errors. Auto-detection before submission. Rejections dropped 80%.
MLETL
BANKING · KYC
7 d → 1 h
Client onboarding
Bank took a week to validate identity. Now 1 hour. Same compliance, 7x more clients onboarded/month.
OCRBiometrics
LOGISTICS · ETL
−46%
Infrastructure costs
Logistics company was paying double on infra due to poorly designed pipelines. Redesign to streaming: −46% costs, real-time data.
AirflowSpark

Six industries, real experience.

I know the domain vocabulary and the systems they use. I don't show up to learn your business on your dime.
Insurance
Claims · Underwriting · Fraud
Healthcare
Billing · Audit · Medical records
Banking
KYC · Scoring · Observability
Retail
Returns · Forecast · Inventory
Legal
Due diligence · Contracts · Search
Logistics
Routing · ETA · Traceability

Stack & expertise.

Production-grade tools I use daily, not things I read about once.
DATA ENGINEERING AI ENGINEERING AUTOMATION MLOPS
Python Airflow dbt Spark / PySpark Databricks Snowflake AWS (S3, Glue, Lambda, Redshift) GCP (BigQuery, Dataflow) Delta Lake Kafka PostgreSQL Docker Terraform CI/CD LLMs (Claude, GPT) Local models (Ollama, Llama, Mistral) RAG MLflow OCR / Document AI Django FastAPI

I publish what others charge for.

Open repos with real production code. If I give this away, imagine what I build when you're paying.
DATABRICKS · ETL
Medallion Pipeline
Bronze → Silver → Gold with Delta Lake, PySpark, data quality and optimization. 10 notebooks.
View repo →
DATABRICKS · DLT
Delta Live Tables
Declarative pipelines with CDC, streaming and built-in data quality.
View repo →
DATABRICKS · ML
MLflow end-to-end
Feature engineering → training → registry → batch inference in production.
View repo →
DATABRICKS · PERFORMANCE
Cost optimization
Skew handling, partitioning, caching, cost management on production clusters.
View repo →

How to get started.

Start with a 30-min call to see if there's a fit. No pitch, no commitment — just a conversation about what's broken and what's possible.
EMBEDDED
Join your team as a senior engineer
Hourly or monthly. Remote, async, timezone-flexible. I plug into your existing workflows — Jira, Slack, GitHub — and start delivering from week 1. No recruiter, no agency cut.
PROJECT-BASED
Scoped engagement, fixed deliverables
Pipeline audit, data migration, automation build. Clear scope, timeline, and cost upfront. Typical engagement: 2–8 weeks. If it doesn't work, we stop.
FRACTIONAL
Part-time senior capacity
Embedded in your data team part-time. Ideal when you need senior capacity without a full-time hire. I bring architecture, best practices, and hands-on delivery.

If your data is unreliable,
your decisions are too.

30-min call. No pitch, no commitment. Just a conversation about what's broken and what's possible.

Connect on LinkedIn WhatsApp