About Cozmo
Cozmo is building AI employees, fully multimodal agents that see, speak, think, and do. We are starting with financial services and insurance, where operations are complex, regulated, and still heavily human-driven.
Our AI employees handle high-stakes workflows with human-level reasoning and communication: real-time customer conversations, claims, underwriting, onboarding, and collections. We are building the infrastructure layer for AI-native banks and insurers, and we want Cozmo to become the operating system for how work gets done in these institutions.
Cozmo is backed by top US investors and is scaling quickly. The ambition is simple and bold: an AI workforce that works as naturally as humans, at a scale humans cannot match.
The Role
You will be one of the earliest senior engineers in the company, and the first wave of leaders in India. This is a role for someone who wants to own real product, not tickets. You will design and build core systems, ship directly into production, and work with founders and customers on problems that do not have a playbook yet. You will be part of the group that future hires call “the original team.”
What You Will Do
- Architect and build real-time multimodal AI systems (voice, text, vision) optimized for ultra low latency and reliability in production.
- Work directly with the founders and senior engineers on product design, technical architecture, and scaling across cloud and edge environments.
- Own end to end pipelines for speech, reasoning, and action, from model integration and orchestration to APIs and infra.
- Collaborate with early enterprise customers to deploy Cozmo AI employees into real workflows and iterate based on live data and usage.
- Help build our India engineering presence by setting high technical standards, mentoring engineers, and shaping how we work as we grow.
You Might Be a Fit If You
- Have built and shipped real products in AI, distributed systems, or real time voice / video. You care about users, not just abstractions.
- Have led projects or small teams before, and enjoy combining high quality hands-on coding with technical leadership.
- Are strong in Python or Rust and comfortable with cloud native architectures (AWS or GCP), containers, and pub/sub systems.
- Have worked with LLMs or multimodal models and like making them fast, efficient, and reliable in real products.
- Want to be early at a company that is trying to define a category, not just optimize an existing one.
Why This Role Is Special
- You get meaningful equity at a stage where it can still move the needle in your life.
- You work with founder level people in AI, product, and finance, across India, the US, and the Middle East.
- You influence the technical and cultural DNA of Cozmo in India. Future leaders will learn from your code and your decisions.
- You get to work on problems that sit at the frontier of LLMs, real time voice, and enterprise scale.
If you want to look back and say “I helped build the AI employee company,” this is that moment.
To apply, send a quick note and your GitHub
🚀 Y Combinator Company Info
Y Combinator Batch: W22
Team Size: 6 employees
Industry: B2B Software and Services
Company Description: Multimodal AI Employees That See, Speak, Think and Do
💰 Compensation
Salary Range: $80,000 - $150,000
Equity Range: 0.2% - 0.8%
📋 Job Details
Job Type: Full-time
Experience Level: 3+ years
Engineering Type: Full stack
Time to Hire: 10
🛠️ Required Skills
Django Go Python Rust
🎯 Interview Process
Our process is fast, founder-led, and focused on real work, not leetcode. We aim to go from first call to decision within 10 days for strong candidates.
Founder Intro (30–45 min)
Goal: Mutual fit check and high-level sell.
What we cover:
- Your story, what you’ve built, and how you think about ownership and risk.
- Cozmo’s mission (AI employees for financial services), current product, and what “founding engineer” really means here.
- High-level tech stack and problem space (real-time voice, agents, infra).
What we’re looking for:
- Founder-level motivation, bias to action, ability to communicate clearly with both technical and non-technical people.
Technical Deep Dive (60–75 min)
Goal: Understand how you build real systems.
Format:
- You walk us through 1–2 real projects you’ve led end-to-end (ideally production systems, AI/infra/real-time).
- We peel into architecture, tradeoffs, scaling, reliability, and how you debugged hard problems.
What we’re looking for:
- Strong architecture thinking, ability to reason from first principles, and pragmatism (shipping vs perfection).
- Experience with LLMs/agents, distributed systems, or real-time voice is a big plus.
Live Build / Pairing Session (90–120 min)
Goal: See how we work together.
Format (collaborative, not adversarial):
- We pick a problem close to Cozmo’s world (e.g., a simplified voice/agent flow, a small service with streaming, or an infra piece).
- You and one of us design and implement a scrappy version together (whiteboard + code).
What we’re looking for:
- How you break down ambiguous problems, choose interfaces, and trade off speed vs quality.
- Your coding style, ownership, and how you communicate while building in real time.
Founder / Culture + Vision Session (45–60 min)
Goal: Decide if we’d be excited to build a company together.
What we cover:
- How you think about startups, intensity, and what “being early” means.
- How you handle conflict, ambiguity, and big technical/product bets.
- Where you want your career and impact to go in the next 5–10 years.
What we’re looking for:
- “Hell yes” founder-fit: someone we’d trust with core architecture, customers, and future hires.