Vondasi

1-on-1 consulting

Tell me about the problem that bothers your team most.

AI comes later. First we look at how often the problem shows up, what it costs you, and what you have already tried. Sometimes the answer is a new app. Sometimes it is one tidied-up spreadsheet.

Services

Everything starts with a 30-minute chat. If it turns out nothing is worth doing, I will say so, and you owe nothing.

Intro Session

30 minutes · Free

For
Anyone with repetitive work, scattered data, or a software subscription that feels too expensive.
Outcome
An honest picture: whether the problem is worth pursuing, and roughly which way to go.
Not included
Digging into company data or detailed technical recommendations.

Software Subscription Audit

Per project · Discussed after the intro session

For
Companies paying per-user software fees while using only half the features.
Outcome
A list of the features actually used, cost per person, the risks if the vendor has trouble, and a build-or-keep calculation that includes upkeep.
Not included
Building the application and negotiating with the vendor.

AI Adoption for Developer Teams

Per project · Discussed after the intro session

For
Software teams that want to use AI coding agents but worry about leaking code or data.
Outcome
Written ground rules (which code may go into AI, which data may not, how to review the output), plus team training.
Not included
Writing code for your team.

Mentoring

1-3 months · Discussed after the intro session

For
Teams building their own systems who need a second pair of eyes for reviews.
Outcome
Regular reviews of architecture and business rules, until the team no longer needs me.
Not included
Taking over the work.

Need the application built as well? That is a separate package. Prices for all services are discussed by email after the intro session.

Contact by email →

How a session works

  1. Fill in a short form: what the problem is, when it last happened, and how you handle it today.
  2. We talk over a video call in your browser, no app to install; the link is in the confirmation email. Most of my questions are about things that already happened, not plans. That is how we see whether the problem is real.
  3. Within 2 business days, you get a one-page written summary.
  4. Whether to continue is your call.

Book 30 minutes, free

Five questions before you build or buy

Free to use on your own. If the answers are unclear, that is good material for an intro session.

  1. Which features are actually used, and what is the real cost per user?
  2. What are the risks of depending on this vendor (lost access, price increases)?
  3. Who will maintain the system once it is built, and at what cost?
  4. What data is stored, and who is allowed to access it?
  5. How will the impact be measured, and when?

Case studies

Every case study states a measured outcome, how it was measured, and its limits.

In-house HRIS replaces a per-employee HR software subscription: an estimated IDR 305-489 million a year in subscription costs avoided

Context
A company in Indonesia with more than 500 employees (anonymised). The project was assigned by the company; I was the developer. The application belongs to the company, so what is shared here is the method and the results.
Problem
The company paid for subscription HR software priced per employee per month, while only using mobile attendance, leave, geofencing, and announcements. Costs grew with headcount, and a late payment once locked the account, disrupting attendance.
Solution
An in-house HRIS covering only the modules in use: mobile attendance, sync with fingerprint machines (check-in and check-out logs only, no biometric data), geofencing, leave, and announcements. The mobile app and the core system are separated: the core system pulls data, and the mobile app cannot reach the core system. Feature code was written with an AI coding agent that I directed and reviewed. I set the architecture and security myself; the UI and the attendance recap logic needed the most corrections from me.
Result
Done in about 1 month by 1 developer (about 66-90 working hours), compared with an estimate of 4-6 months for 2 developers on a similar internal project without AI. In full use since June 2026. Estimated subscription cost avoided: IDR 25.5-40.7 million a month, or about IDR 305-489 million a year: headcount times a per-head rate of IDR 50-80 thousand, not an invoice figure.
Limits
The savings figure is a gross estimate: server, maintenance, and staff time are not deducted. The comparison project's scope is not identical. Operational impact (HR recap time, manual corrections) has not been measured yet. Payroll, social security, and tax document distribution are not in the application yet, so the claim only covers the modules replaced.

About me

Panji M. Al Latief, Senior IT Application Developer, Bekasi, Indonesia

I am Panji, a developer in Bekasi. Day to day I build internal applications: attendance, assets, helpdesk. This year I built an HRIS for more than 500 employees, largely with an AI coding agent, in about a month. It taught me which parts AI really speeds up, and which parts still need someone who understands the process.

Skills

Web
Go, Java Spring Boot, HTML, CSS, JavaScript, Tailwind CSS
Cloud and infrastructure
Docker, Proxmox, Google Cloud (GCP)
AI
LLM integration, Agentic AI, RAG (foundational)

Project experience

Described without system or client names to respect confidentiality.

  • Employee attendance system connected to fingerprint devices and a mobile app.
  • Asset management system with role-based access control (RBAC).
  • Data ingest and integration platform for many contributors.
  • IT service platform: helpdesk, KPI tracking, and QR code asset tagging.
  • WhatsApp bot for finance record keeping.
  • RFID-based e-canteen and point-of-sale (POS) app.

Education

  • M.Kom (Master of Computer Science), Informatics

    President University, Cikarang, Indonesia · Graduated January 2025

    Thesis: Enhancing Plant Recommendation through IoT-Integrated LLM Systems

  • S.T (Bachelor of Engineering), Information Technology

    Gunadarma University, Bekasi, Indonesia · Graduated August 2021

    Thesis: Measurement Tool for Planting Medium Using NodeMCU Based Android

Contact

Not sure your problem fits? Send a two- or three-sentence email first. I will reply and tell you honestly whether a 30-minute session is worth it.