Software Engineer

We’re hiring a
Software Engineer
to help build the core AI systems of an early-stage company. You’ll work directly with the founders to design and ship
production-grade, agentic AI workflows
that solve real customer problems end to end.

This is
not a research role or model-tuning role
. The focus is on building
reliable, scalable software
: systems where AI plans, executes, verifies work, and integrates cleanly with existing tools and data. You’ll have
high ownership from day one
shaping core architecture, influencing AI agent behaviour, and helping define best practices for how modern AI-native software is built.

What You’ll Do

- Design and ship
agentic AI systems
that operate in real, messy environments
- Build
production workflows
that combine LLMs, deterministic logic, and human-in-the-loop patterns
- Own features end-to-end—from problem definition to deployment
- Work closely with founders and early users, iterating quickly based on feedback
- Make
architectural decisions
that scale from early customers to large deployments

What We’re Looking For

- 1–5 years of professional software engineering experience
- Experience at
early-stage hypergrowth startups
(Seed to Series B)
- Hands-on experience
building production-grade AI agents and workflow systems
- Previous founder experience or
exceptional extracurricular achievements

Hard Skills

- Proficiency in
Python
and core data science libraries
- Proficiency in
Next.js
for frontend development
- Experience with
AWS
for cloud-native infrastructure
- Understanding of
GPU infrastructure and model optimization
- Experience with
LLM orchestration frameworks
(e.g., LangChain, LlamaIndex)
- Skilled at
optimizing small language models for scalability
- Clear
design sense & taste for good UX/UI

Soft Skills

• High ownership, strong product intuition, and comfort operating in ambiguity
• Ability to iterate quickly and make architectural decisions

Nice to Have

- Experience
scaling ML models in production environments
- Experience designing
AI workflows beyond simple prompt-response