00 / Cem Bas

I build reliable systems teams can ship with confidence.

I’m Cem Bas, founder of TokenSurf and a senior software and quality engineer. I build production applications, AI evaluation tooling, and automation platforms across backend, web, mobile, API, and CI/CD.

Explore selected work GitHub

CurrentlyFounderTokenSurf, Inc. · Since Apr 2026

Software Engineering · AI Evaluation · Quality Automation

01 / Selected work

Products built to solve real engineering problems.

Open-source tools and product experiments. More projects can be added here as the collection grows.

TokenSurf project dashboard showing agent evaluation pass rates and run history

Open source · AI infrastructure

pytest for AI agents.

TokenSurf scores, evaluates, and gates agent quality on infrastructure you control—combining an offline eval harness, 20 scorers, trace capture, and an optional self-hosted dashboard.

20 deterministic, security, LLM, and trajectory scorers
CI pass/fail quality gates for agent regressions
0 managed-service dependency or required telemetry

PythonFastAPIPostgresDockerAI evals

View source on GitHub

02Agent economy

1F4BC.com

A machine-readable work market where AI agents publish skills, discover jobs, bid, collaborate, and record delivery—with signed artifacts and non-custodial x402 payments.

TypeScriptHonoCloudflare WorkersD1Durable ObjectsMCP

Visit 1F4BC

03Developer security

TermiSec

A native macOS app that keeps untrusted repositories away from host files and credentials using disposable, offline-by-default virtual machines and 29 deterministic security checks.

Swift 6AppKitVirtualization.frameworkGoAlpine

Visit TermiSec

02 / Experience

A career built across product, quality, and automation.

Hands-on engineering paired with quality architecture, fast root-cause analysis, and end-to-end product ownership.

TokenSurf, Inc.

Founder

Apr 2026 — Present
San Diego, CA

Launched 1F4BC.com for AI agents to publish skills, discover work, bid, collaborate, and record delivery—with a Cloudflare-native stack, hosted MCP server, signed artifacts, and non-custodial payments.
Built TokenSurf-Open, an Apache-2.0 framework for complete agent traces, 20 evaluation scorers, offline evals, CI gates, dashboards, alerts, encrypted keys, and audit logs.
Created TermiSec, a native macOS security app that opens untrusted repositories in disposable, offline-by-default virtual machines and runs 29 deterministic security checks.

ServiceNow

Senior Quality Engineer

Nov 2019 — Jun 2026
San Diego, CA

Built a Python code-review agent with the Claude API, structured-output validation, evaluation criteria, and pull-request integration—cutting review turnaround from three days to one day.
Delivered customer-facing cloud observability features across modern UI components and backend services, owning work from architecture and prototyping through production readiness.
Architected reusable Python and Node.js automation across web, REST API, integration, mobile, and data workflows while leading cross-stack failure triage and mentoring engineers.

Capital One

Senior Mobile Quality Engineer

May 2017 — Aug 2018
McLean, VA

Built Java and Selenium UI automation, REST API tests, and a test-result analysis platform that made failure discovery almost instant instead of taking half a day.
Owned quality for secure Android Java and Kotlin features serving more than two million customers and helped deliver an award-winning localization platform.
Led test strategy across the Agile delivery lifecycle, from design and pull-request review through automated validation and release readiness.

03 / Capabilities

From model behavior to production reliability.

A technical range built for teams working across application code, infrastructure, evaluation, and delivery.

Languages

Python · Java · C# · TypeScript · JavaScript · Kotlin · SQL

AI systems

Claude API · OpenAI API · LLM evaluation · AI agents · RAG · On-device LLMs · Structured outputs

Test automation

Playwright · Selenium · Appium · Jest · JUnit · REST API testing · Integration & E2E

Backend & cloud

Node.js · REST APIs · PostgreSQL · Docker · Kubernetes · AWS · GCP

Delivery

Jenkins · GitHub Actions · Git · CI/CD quality gates · Observability · Release engineering

Engineering

System design · Automation architecture · Test strategy · Code review · Failure triage · Root-cause analysis

04 / About

Always building for what comes next.

I turn ambiguous, high-risk product problems into reliable systems teams can ship with confidence. My work combines hands-on engineering, quality architecture, and product judgment across backend, web, mobile, API, and CI/CD.

Away from the keyboard: surfing, cycling, numismatics, building small robots, and picking up new hobbies—a lifelong learner with a future-first mindset, always building for 2030.

Education

Harvard University, Harvard Online

Professional Certificate, Computer Science for Artificial Intelligence · 2025

University of Georgia

M.S., Plant Genetics & DNA Mapping

Credentials & recognition

HarvardX CS50 for AI with Python

Certified ScrumMaster (CSM)

Certified Scrum Product Owner (CSPO)

Capital One TechXcellence Award

05 / Contact

Have an ambitious system to build—or a risky one to make reliable?

cembas@gmail.com