andrei.sas
OPEN
[ session // 2026.05 ]● open to opportunities

[ ANDREI SAS — file: index.tsx — line 001 ]

ANDREISAS_

AI-first software engineer. I ship production systems — front to back — in days, not months. Worked with Google and Deloitte via Baseflow (NL). Now building under Aumovo. Open to roles, contracts, and partnerships — remote, anywhere.

[ selected output↓ ][ get in touch→ ]
terminal // boot.logtty/01
>sys init / connecting agent...
>auth OK / handshake @ 2026.05
>loading aumovo/andrei.sas...
>providers: anthropic · openai · vertex
>
runtime
claude-code
latency
~24h reply
stance
ai-native
basis
remote · global
[ 01 ]5+yrs engineering
[ 02 ]$200K+client revenue / 1 campaign
[ 03 ]60%cold-email open rate
[ 04 ]04industries shipped
[ collaborated with ]
BASEFLOWGOOGLEDELOITTE+ B2B CLIENTSBASEFLOWGOOGLEDELOITTE+ B2B CLIENTSBASEFLOWGOOGLEDELOITTE+ B2B CLIENTS
ai-first·ship in days·not months·studio-grade creative·claude code native·vertex agents·ai-first·ship in days·not months·studio-grade creative·claude code native·vertex agents·ai-first·ship in days·not months·studio-grade creative·claude code native·vertex agents·
[01/ 05]

client engagements & employment

[ NDA-bound engagements omitted — happy to walk through under nda ]

[02/ 05]

shipped solo / live or in build

60% open rate//$200k+ generated//3,000+ prompts shipped//5+ yrs eng//ex-google × deloitte//cinematic ai//60% open rate//$200k+ generated//3,000+ prompts shipped//5+ yrs eng//ex-google × deloitte//cinematic ai//60% open rate//$200k+ generated//3,000+ prompts shipped//5+ yrs eng//ex-google × deloitte//cinematic ai//
[03/ 05]

how I think / why it ships

I started as a normal software engineer. Five years of writing it by hand. Native apps, multi-tenant platforms, distributed systems — the kind of work that used to need a team and a quarter.

Then AI happened, and I retooled from the ground up. Most engineers bolted Copilot onto an unchanged workflow. I tore the workflow apart and rebuilt it around agents — Claude Code, prompt pipelines, custom retrieval, deterministic guardrails, agents deployed on Google Vertex.

The result: I ship in days what used to take months. A custom CRM, lead-gen pipeline, and outbound system — week. An emissions-management platform with AI invoice parsing, role-based reporting, and live support — a month. Production quality. Not prototypes.

I'm open to senior engineering roles, contracts, and partnerships — anywhere remote. If you're trying to put 10 engineers' worth of output behind one person, that's the conversation.

[01]
Ship the whole thing.
Frontend, backend, infra, agents, prompts, docs. One operator, one delivery.
[02]
Optimize for revenue.
Pretty doesn't matter if it doesn't earn. Every feature traces to a number.
[03]
Measure twice. Automate.
If you do it more than three times, the system does it next time.
[04]
Trust the model. Verify the output.
AI-first doesn't mean AI-blind. Every loop has a checkpoint.
// timeline2014 → now
  • 2020
    started shipping software
    first production code. learned the long way — before AI made it short.
  • 2021
    baseflow / NL
    joined a top-tier dutch software firm. maintained open-source flutter packages used by millions.
  • 2022
    google × deloitte
    worked directly with the google flutter team in the US. shipped enterprise tooling for deloitte.
  • 2023
    went ai-native
    stopped writing code by hand. rebuilt my entire workflow around llms. started shipping 10× faster.
  • 2024
    built aumovo
    started taking on senior engineering work directly — ai platforms, custom CRMs, internal tooling.
  • now
    open to what's next
    engineering systems that earn — open to senior roles, contracts, and partnerships, remote anywhere.
[04/ 05]

what I work with / how I work

[ 01 ]languages
  • +TypeScript
  • +Python
  • +Dart
  • +Go
  • +SQL
[ 02 ]frameworks
  • +Next.js
  • +React
  • +Flutter
  • +FastAPI
  • +Node.js
[ 03 ]ai providers
  • +Anthropic
  • +OpenAI
  • +Google Vertex
  • +RAG / vector search
  • +Agents & tools
  • +Prompt engineering
[ 04 ]data
  • +Postgres
  • +Supabase
  • +Redis
  • +Pinecone
  • +Drizzle / Prisma
[ 05 ]infra
  • +Vercel
  • +Google Cloud / Vertex
  • +AWS
  • +Docker
  • +GitHub Actions
  • +Edge runtimes
[ 06 ]practice
  • +Claude Code
  • +Agentic dev
  • +Outbound automation
  • +Lead enrichment
  • +Email infra

how I work.

  1. 01
    Understand the system
    Before code, I map the business, the constraint, and the revenue lever. No deck — just questions.
  2. 02
    Ship in slices
    Working software over status updates. Daily checkpoints, branch-shipped features, no surprise reveals.
  3. 03
    Trust the model. Verify.
    Agents and LLMs do the typing. I own the design, the guardrails, and the proof it actually works.
  4. 04
    Document and own outcomes
    Whether I'm an FT senior eng, a contractor, or a partner — I'm measured on what shipped, not hours logged.
[05/ 05]

reply within ~24h

Hiring, building, or just exploring — say hi. Open to senior engineering roles, contracts, and partnerships. The more concrete the better — a role you're filling, a system you're building, a workflow you'd kill to automate.

// what fits
  • +Senior eng roles — full-time, fractional, advisory
  • +AI platforms (RAG, agents, document AI, Vertex)
  • +Custom CRMs / internal tooling
  • +Outbound + lead-gen systems
  • +Greenfield SaaS — front to back
// directonline
andrei@aumovo.com↗
basedbucharest · remote · global
operating asaumovo
/contact/newfields: 04 · req: 03

[ end-to-end / no tracking / I read every message ]