GatherInsight: research data that comes with its proof.
Surveys, interviews, traffic counts and geo mapping. One web platform, three mobile apps.

Full Stack
I build data products, end to end. From ML models and dashboards to Next.js SaaS, mobile field apps and bots that do the busywork.
AI that sees and predicts, live dashboards, apps and bots. Open any of them and play.
( 01 )What I do
Idon'tstopattheinsight.Ibuildthewholething:theappthatcollectsthedata,themodelthatreadsit,andtheautomationthatactsonit.
Led data work at
( 02 )What I build
Products, portals and internal tools that people log into every day.
( 03 )Selected work
Products, apps and bots that left the notebook and went to work.
Surveys, interviews, traffic counts and geo mapping. One web platform, three mobile apps.

The field app Indosat promoters work in: check in at the outlet, record each SIM card sold, see whether it activated.



The daily workbench of Indosat's Kalisumapa channel team: partners, distribution, payouts, revenue and a shelf of tools behind one login.

The helpdesk behind the channel team: areas raise a ticket, a handler picks it up, and nothing goes quiet without a trace.

Data owners paste an Excel for the web embed once, the admin files it under a menu, and viewers open it without hunting for links.

One place for the circle, its regions and the staffing agencies to see every sales slot: who's in, who left, and how long a seat has stood empty.

Marketing's spending in one view: what was approved, what's committed, what's been paid, and what's left each quarter.

Staff log what they buy for each event with a photo of the receipt; admins review it, and the totals add up on their own.

A web app replaced the Oracle Fusion sales-order macro: upload the quantity sheet, every row is checked, and the import files come out zipped. Plus bots for Mobii, SalesHub and Mobile Money.

A D-1 KPI wall for the office TV: every region, area and branch on one screen, rebuilt from the morning's Excel exports. Plus GA funnels, partner quadrants and fraud monitoring on the web.

From thesis to a live dashboard stakeholders check in real time.
( 04 )How I ship
The same four steps, whatever I'm building.
Same flow for
Listen
Prototype
Ship
Keep improving
Listen
( 05 )Journey
From media analytics to telco operations. Keep scrolling →
2026 — Now
Mitracomm × Indosat
Internal portals, dashboards and bots for Sales & Distribution.
2024 — 2026
Telkomsel Distributor
Databases, web dashboards, fraud detection.
2023
Kalla Group
Pipelines and price forecasting.
2023
Bangkit Academy
ML, deep learning, NLP, vision.
2020 — 2024
Universitas Pancasila
Data Science · GPA 3.76
2021 — 2022
INIPASTI Communika
Audience analytics for media.
( 06 )Toolkit
drag the dots around ↓
Plus 20 experiments in ML & deep learning
( 07 )Explore my skills
Each skill has its own page: how I work, the projects, and demos you can try.
Turning history into predictions people can act on.
17 case studies with models you can run
Neural networks for images, text and time series.
13 demos: camera, drawing, forecasts
Questions answered with SQL, charts and dashboards.
279 charts · 4 live dashboards · data models
Pipelines that make messy data trustworthy.
Pipeline lab · incident simulator · 6 architectures
Products people log into every day.
7 pages · SaaS sandbox · mobile lab · concepts
Bots and agents for the repetitive work.
7 pages · workflow lab · AI agents · prompt → automation
( 08 )About
0+
apps, dashboards & bots
0
portal automations
0%
decisions enhanced
0.00
GPA
2020 — 2024 · Jakarta
Universitas Pancasila · GPA 3.76 · Thesis on RNN forecasting
Database Administrator
BNSP · valid to 2027
TensorFlow Developer
Google · valid to 2026
Bilingual
English · TOEFL ITP 617 · IELTS 6.0
Always learning
Based in Makassar, working across Indonesia.
( 09 )Contact