PELOPS² — Founding Software Engineer (Backend / Full‑Stack)

Build the intelligence that doesn’t just suggest the next step—it finishes the work. We’re creating agentic AI that runs real operations end‑to‑end for small businesses.

The Problem

Today’s AI is great at drafts and suggestions—but it rarely closes the loop. Small businesses still lose hours every week chasing clients, scheduling meetings, coordinating vendors, and following up on invoices. The “last mile” of execution is missing.

The Opportunity

We’re starting with professional virtual assistants in the Philippines—the people who already run day‑to‑day operations for many U.S. SMEs across sales, marketing, customer service, admin, and specialized roles like real estate and healthcare. The Philippines is a leading hub for skilled, English‑proficient talent supporting global businesses, making it an ideal proving ground for real‑world workflows.[1]

What We’re Building

PELOPS² is an execution‑first AI: it senses context, plans, acts across tools, learns from feedback, and collaborates—so tasks get done, not just drafted. We’re designing the memory and reasoning layer that helps humans and AI work together day‑to‑day, with agents that can prioritize, comply with guardrails, and know when to loop in a human.[2][3]

Where We Are

• Pre–product‑market fit with strong problem validation.
• Moving fast toward our first public release (Version 1 ships October 2025).
• Early stage where decisions matter, impact is immediate, and everything you build becomes part of the foundation.

The Role

You’ll be our first (or one of the first) backend/full‑stack engineers. You’ll own systems end‑to‑end—from designing the data model and agent workflows to shipping production features, observing them in the wild, and iterating quickly.

What You’ll Do

• Design and build event‑driven backends that orchestrate multi‑step workflows across CRMs, calendars, email, messaging, and payments.
• Implement memory, reasoning, and guardrails so agents can prioritize, execute safely, and close tasks autonomously.
• Ship pragmatic LLM‑powered features (prompting, RAG, tool‑use, summarization, reconciliation) with measurable business outcomes.
• Integrate third‑party APIs (e.g., scheduling, billing, telephony) and create internal SDKs for repeatable tool integrations.
• Stand up robust observability (traces, metrics, evals) to diagnose and improve agent behavior in production.
• Partner directly with users (expert VAs) to encode real workflows and reduce steps to outcomes.

What You’ll Bring

• Experience building and owning systems end‑to‑end (bonus if in Python) with a willingness to ship, break, and fix things responsibly.
• Product sense: strong bias for simplicity, data‑driven iteration, and crisp user outcomes.
• Systems thinking: you see interactions, not isolated features; you design for failure and feedback.
• Comfort with ambiguity and a builder’s mindset; you value craftsmanship and clarity.

Bonus Points

• Hands‑on work with LLMs, workflow engines, or agentic architectures (planning, tool‑use, multi‑agent orchestration).[4]
• Experience integrating business systems (CRM, ERP, support, billing) and building internal platforms/SDKs.
• Prior startup or founding‑team experience.

How We Work

• Ownership over location and schedule; what matters is outcomes and reliability.
• Write it down: crisp specs, short design docs, and lightweight runbooks.
• Ship small, measure, iterate—favor reversible decisions and high‑leverage tools.
• Respect for users’ time: observe, learn, and reduce work for them every release.

Compensation & Upside

• Equity‑first (founding‑team level), designed for long‑term upside and high trust.
• Remote‑friendly—location doesn’t matter; ownership does.
• Wide creative latitude with the expectation of high standards.

Why Now (and Why Start with VAs?)

Virtual assistants already operate as the execution layer for many SMEs; they’re embedded in daily business systems and processes. The Philippines, in particular, is a premier hub for trained, English‑proficient talent serving global clients—making it an ideal environment to ground, evaluate, and scale real operational workflows.[1]

The Name: Why “PELOPS²”

Inspired by Godzilla: Singular Point, where Pelops II is a personable AI companion that collaborates closely with a human researcher—distinct yet inseparable. That spirit of human‑AI partnership is core to what we’re building.[5]

Tech You’ll Touch (representative, not prescriptive)

• Python, FastAPI, Postgres; event queues and schedulers for orchestration.
• LLM orchestration, retrieval (RAG), vector stores, tool‑use, evaluation harnesses.
• API integrations for calendars, CRMs, email, chat, telephony, and payments.
• Observability: tracing, metrics, red‑teaming, and feedback loops.

What Success Looks Like

• Tasks move from “drafted” to “done” with minimal human intervention.
• Users report fewer steps to outcomes and higher on‑time completion rates.
• Clear paper trails and guardrails keep execution safe and auditable.
• The system gets smarter with use—learning from feedback and outcomes over time.

How to Get in Touch

• Send a brief note with links to something you shipped (what you built, what broke, and what you learned).
• If you have examples of workflow automation or LLM‑powered features, we’d love to see them.

Let’s build the execution layer for work—an assistant that anticipates, acts, and adapts like a true teammate.

About PELOPS²

The project is developing an AI-first system that learns how users work, remembers context, manages tasks, and keeps daily operations in perfect flow—essentially “auto-complete for work” that goes beyond simple suggestions to full execution.

Vision: AI That Works Like You Do

Most business operations today rely on software demanding manual clicks and rigid rules. The new AI system aims to transform this by learning actual business workflows and completing jobs end-to-end, minimizing manual steps and human intervention.

Real-World Foundation: Philippine Virtual Assistants

The initiative starts with virtual assistants in the Philippines—professionals who manage all aspects of daily business for U.S. owners:

• Sales and marketing
• Customer service
• Administration
• Specialized support in real estate, healthcare, and more

These VAs are vital to how many American SMEs truly operate day-to-day and provide a rich training ground for the system’s AI.

Why Train on VAs First?

• Their workflows are complex, varied, and deeply embedded across business functions.
• Capturing the way VAs identify needs, prioritize, and execute tasks gives the AI practical, adaptive skills for real-world business automation.
• This builds a strong foundation to later serve SMEs and service teams in the U.S.—segments overlooked by most AI automation today.

End-to-End Automation and Impact

By learning directly from these workflows, the AI system will:

• Anticipate context and requirements
• Manage and execute tasks from start to finish
• Streamline operations and scale easily across business teams

Ultimately, this approach moves AI from a supporting or drafting role to a true operational teammate—capable of handling business activities with intelligence, adaptability, and minimal supervision.

The company culture emphasizes outcome-driven work, respect for global perspectives, and deep individual ownership in a flexible, remote-first environment. Below is a beautified, unique culture overview for your startup:

Culture Overview

Focus on Outcomes, Not Optics

• Measured by what is built, not hours spent at a desk.
• Craftsmanship and clarity matter more than politics or physical presence.
• Autonomy and trust are foundational—results make the biggest impact.

Global Talent, Shared Mission

• Remote-first lets us recruit the best talent, no matter where they live.
• A team that spans time zones brings diverse perspectives and rich ideas.
• Unified not by geography, but by a shared mission to build transformative technology.

Deep Work > Constant Noise

• Avoid interruptions common to the traditional office; remote culture favors deep, focused thinking.
• Async communication is optimized—fewer meetings, more time for breakthroughs.
• Every teammate has space for clarity, innovation, and flow.

Ownership from Day One

• Teammates own their work conceptually and creatively—no middle layer or bureaucracy.
• Define architecture, shape product, and deploy code rapidly for direct impact.
• Everyone is empowered to challenge, improve, and ship at speed.

Work Built Around Life

• Flexibility is a core belief: work adapts to life, not the other way around.
• Trust teammates to build their own professional rhythm, wherever they are.
• Great outcomes thrive when people are free to live deeply and work meaningfully.

What Makes Us Unique

• Radical transparency in workflow, decision-making, and goal-setting.
• Culture of empathy, asynchronous collaboration, and self-driven mastery.
• Remote-first isn’t a trend; it’s how the best teams create, learn, and build together.

This remote-first model empowers every individual to contribute, transform, and thrive—united by impact, shared vision, and the freedom to do work that matters regardless of location.