AI-native operations
I run my own work on PM-OS, an operating system I designed and shipped: meeting capture, semantic knowledge base, daily briefs. Built, not bought.
PM-OS / SELF-HOSTED / IN DAILY USE
Rama
Syaliandra
LEADS PRODUCT. BUILDS SYSTEMS. TRADES IDX.
8+ years shipping Indonesian fintech, from IDR 1.6T lending rails to AI systems and systematic IDX trading I build myself.
syaliandra.dev
zsh
8+ YRSPRODUCT LEADERSHIP
IDR 1.6TMONTHLY DISBURSEMENT
-15%NPL RATIO
500+PMS MENTORED
350K+AGENTS SERVED
2,000+VEHICLES ONBOARDED
Now
I am a product leader operating in Indonesian fintech. My last public chapter: Head of Product at AwanTunai, scaling MSME lending to IDR 1.6T in monthly disbursement with a product organisation of nine.
What makes me different is that I build. I trade IDX with Vantage, a screener I built from scratch, and I run my own work on an AI operating system I designed and shipped. The judgment comes from leading. The edge comes from building.
If you are scaling a product organisation, that combination is the point: strategy set at the top, verified hands-on.
JAKARTA, INDONESIA
Quant trading on IDX
I trade Indonesian equities with Vantage, a decision-support screener I built: broker-flow attribution, intraday signal scoring, and AI session summaries.
VANTAGE / LIVE CAPITAL / IN DAILY USE
Growing product people
Led a product org of 9 at AwanTunai. Lead Lecturer at RevoU, where 500+ aspiring PMs across Indonesia went through curriculum I designed.
ORG OF 9 / 500+ MENTEES
Journey
Current
01 / 11
WORK://
Lead Product Manager at Undisclosed, Indonesian Fintech
Operating in a lead product role at an Indonesian fintech. Details when the time is right.
02 / 11
FOUND://
Builder & Operator at Vantage, IDX Decision-Support Screener
Spent a deliberate year between roles building Vantage, a decision-support screener for Indonesian equities, and PM-OS, my personal AI operating system. Both run in production and I use them daily.
67Klabeled alerts behind scoring v25phases shipped to production
AwanTunai
03 / 11
WORK://
Head of Product at AwanTunai, Fintech Lending
Scaled the lending portfolio across microloans, supplier financing, and trade financing. Launched ML risk scoring that cut approval time from days to under 24 hours, and built an Early Warning System that lowered the NPL ratio. Led a product organisation of 9.
IDR 1.6Tmonthly disbursement-15%NPL ratio+35%loan outstanding in 2 quarters
04 / 11
WORK://
Group Product Manager at AwanTunai
Defined long-term product strategy and half-year OKRs with a 90% delivery rate. Redesigned the SDLC, shortening the delivery cycle by roughly 20%.
05 / 11
WORK://
Senior Product Manager at AwanTunai
Designed the roadmap that expanded MSME lending penetration by about 25% YoY, grounded in 50+ merchant interviews and discovery sessions.
RevoU
06 / 11
EDU://
Lead Lecturer, Product Management at RevoU, Edtech
Led the PM lecturer team, redesigned the curriculum around current fintech and SaaS practice, and mentored 500+ aspiring product managers across Indonesia.
500+PMs mentored
07 / 11
EDU://
Lecturer, Product Management at RevoU
Taught Product Vision and Strategy, Discovery, and Roadmapping fundamentals while contributing to curriculum development.
Earlier
08 / 11
WORK://
Product Manager at PayFazz (now Fazz), Agent Banking
Owned the roadmap for an agent banking network serving 350K+ agents. Ran 100+ customer studies that lifted retention 20% and NPS by 15 points.
350K+agents nationwide
09 / 11
WORK://
Product Manager at Dukung Calonmu, Political Tech
Built and scaled a campaign fundraising platform supporting 500+ candidates, raising donor engagement 40% through payment integration and real-time tracking.
10 / 11
FOUND://
Co-Founder & Product Lead at Garazi, P2P Car Rental
Founded a peer-to-peer car rental marketplace from zero, onboarding 2,000+ vehicles. Secured seed funding and grew active users 60% YoY before a strategic wind-down.
2,000+vehicles onboarded
11 / 11
WORK://
System Analyst Intern at Microsoft Indonesia
Structured business requirements for 3 cross-department projects and ran system workflow training that cut repetitive support queries by about 20%.
Recognition
- 2019Telkomsel NextDev Winner, NextDev Palembang
- 2019Beehive Best Performing & Most Innovative
- 2014GEMASTIK 7 1st Place, Software Development
- 2013GEMASTIK 6 2nd Place, Game Development
- 2013MGDW 4 Best Cultural App
Builds
01
LIVE
2025 - now
Vantage
A decision-support screener for Indonesian equities. It watches the whole market, attributes broker flow, and scores intraday signals so the trader makes the final call. Not a bot, by design.
The situation
IDX has no retail API, and reading smart-money flow is manual, slow, and expensive. I wanted institutional-grade flow analysis on a retail trader's desk, starting with my own.
The calls I made
- Built the entire data layer from scratch: end-of-day tape, broker summaries, and a 3-second orderbook feed.
- Drew the product line at decision support, not automation. Vantage surfaces evidence and conviction tiers, the trader makes every call.
- Rebuilt the scoring engine from data, not intuition: per-alert lift analysis over 67,000 labeled alerts set the v2 weights, and every decision is stamped with a weights version for post-deploy monitoring.
- Encoded market microstructure as guardrails: tick-size tiers, special-monitoring-board filters, and an optional Sharia mode.
67Kalerts labeled for scoring v2~900stocks scanned every 60s3sorderbook resolution
TypeScript / Bun / Next.js / Fastify / PostgreSQL / Claude
02
LIVE
2026 - now
PM-OS
A personal operating system for product leadership: meeting capture, task clarification, a semantic knowledge base, and daily briefs. Written as a PRD first, then shipped through gated slices.
The situation
Product leadership runs on operational overhead: notes, tasks, email, and context scattered across tools that never share what they know. I wanted one system where every interaction feeds a knowledge base that compounds.
The calls I made
- Wrote the PRD before the code: baseline-before-optimizing metrics, explicit non-goals, and pass/fail gates for every slice.
- Designed one shared backbone (a PostgreSQL and pgvector knowledge store plus an LLM layer) with capture decoupled from processing, so the core survives any change of device or tooling.
- Made privacy an architecture decision: self-hosted transcription, local embeddings, passkey auth, encrypted tokens, and a confirm queue so nothing is sent without sign-off.
- Shipped slice 1 end to end: capture to clarify to schedule to calendar, as an installable PWA.
10/10slice-1 steps shipped94tests green100%self-hosted pipeline
Bun / TypeScript / PostgreSQL / pgvector / Claude / Whisper
03
SHIPPED
2020 - 2025
AwanTunai Lending Suite
Five years of end-to-end product leadership for MSME lending in Indonesia, from senior PM to Head of Product, across microloans, supplier financing, and trade financing.
The situation
Indonesia's MSMEs run on informal credit. AwanTunai lends against supply-chain data that banks cannot see, which means the product is the risk engine.
The calls I made
- Built and led a product organisation of 9 across lending, risk, and repayment.
- Bet on ML risk scoring over manual review: approval time fell from days to under 24 hours, and loan outstanding rose ~35% in two quarters.
- Drove virtual-account repayment adoption from under 10% to 40%, lifting repayment efficiency by ~25%.
- Built an Early Warning System that cut the NPL ratio by 15% while the portfolio kept scaling.
IDR 1.6Tmonthly disbursement-15%NPL ratio<24happroval time, from days
B2B Lending / ML Risk / Team of 9 / 4 Product Lines
04
ARCHIVE
2016 - 2018
Garazi
Peer-to-peer car rental marketplace built from zero to seed funding as technical co-founder, covering product, design, and engineering.
2,000+ vehicles, seed funded
Marketplace / Product / Engineering
Let's compare notes
Scaling a fintech or AI product organisation in Indonesia or Southeast Asia? I am open to product leadership conversations, advisory, and quant research collaboration. I usually reply within 48 hours.