Use or Space
Probation Confirmation Report · 2026 转正述职报告

Building the technical backbone
for overseas client.

Rizky Azmi Swandy
Technical Presales Engineer · Forward Deployed Engineer
01 — Work Review · 工作回顾

Technical ownership in Jakarta team

01
Technical validation & R&D
Benchmarking, daily model testing, and knowing our platform's real performance.
02
Forward-deployed engineering
On the ground with clients' engineers to resolve issues and help them use Nebula effectively.
03
The bridge to HQ
Centralized, evidence-based tracking and escalation of platform issues.

In a small Indonesia team, every role has to carry real weight. I naturally focused mine on reducing technical uncertainty for clients, sales, and HQ.

02 — Results · The technical layer

Testing, tracking, resolving

Keluh — my experimental lab
Model testing and live issue handling, every day
Direct gateway access + prod-DB debugging
Traced a billing issue at the source — before the backend update
Centralized issue tracking
24 (and counting) issues logged with reproduction and evidence
Baseline latency — tracked in my regular R&D
Nebula9,619 ms
OpenRouter4,978 ms

Same model, 39,946-token input. Run regularly so we always know where our platform stands.

02 — Results · Forward deployed

On the ground with clients' engineers

MiminOn-site

Deployed to client's office to investigate Nebula API implementation issue with their engineers — isolated the root cause and helped turn doubt into trust. Impact: Mimin is preparing to migrate 15 live client projects to Nebula API, including Salim Group's Bank INA and DMC Hospital, plus 10 staging projects targeted to go live next month.

SingtelOnline

Paired with their engineering team on integration and DX, guiding effective Nebula API use. After two sessions: zero issues reported — all good on their side.

Pairing with client engineers — the work I do best.

02 — Results · Before the room

The technical work behind every meeting

Pre-meeting technical briefs and client research, so the team walks in prepared — plus replicating a client's own audit script to validate Nebula end to end.

MiminTechbridgeRuangGuruTelkom STOTelkom DPEDisprzParateknoSingtel Lintasarta

Working directly with Indonesia and overseas clients.

03 — Shortcomings & Countermeasures · 不足与对策

What's hard — and what I'm doing about it

Some platform issues are still ongoing and need to be tackled from the root. When fixes sit in the codebase, release cycle, or HQ roadmap, my local role is to investigate deeply, narrow the cause, and make escalation evidence-based.
Countermeasure — centralized, evidence-first tracking · Keluh self-serve testing · direct debugging within my access · and a regular weekly sync with the China team to move cases from issue → root cause → fix → learning, not only report → wait.
Platform upgrades aren't yet surfaced to the Jakarta branch — we're not notified when changes ship, or what they cover, and the history isn't documented in a digestible form, so we can't anticipate client-facing behavior shifts.
Countermeasure — I monitor the codebase directly and run Keluh regression checks to catch drift early — and propose lightweight release notes alongside the weekly sync, so changes surface before they reach clients.
Observability is limited — issues are hard to trace end to end without a correlation id.
Countermeasure — pushed for correlation ids, better logging, and a concrete metadata backlog so issues can be traced from client request to platform behavior.
04 — Development Plan · 发展规划

Where I can add the most value

Overseas platform engineering
Handle the overseas platform layer within my access: reproduce, debug, validate, and deploy fixes myself where possible — then escalate deeper changes with technical context.
Faster issue ownership
Move from report-and-wait to investigate-and-sync: solve what can be solved locally, then align HQ with evidence for deeper changes.
HQ feedback loop
Weekly sync with the China team so overseas issues become platform improvements, not repeated support tickets.

The goal is to make Indonesia's technical support more engineering-led: faster local handling, clearer HQ handoff, and less uncertainty for clients.

04 — Development Plan · Growth

How I grow into engineering ownership

Career direction
  • Move deeper toward engineering, not only technical support
  • Build stronger platform and codebase understanding
What changed
  • Client deployment shows real integration problems directly
  • Codebase access since last month helps me debug beyond surface symptoms
60 – 90 days
  • Formalize the overseas platform issue loop
  • Weekly HQ sync with root-cause and solution notes
  • Deepen SEA client engineering support

Client deployment and codebase access changed how I work: I can debug real causes, not only escalate symptoms. That is the direction I want to grow — deeper engineering ownership for overseas clients.

Thank you

一段旅程结束,另一段征程又开始。
The journey of grow with Nebula — Rizky Azmi Swandy