Les missions du poste
Important information
Contract type: Permanent contract
Salary: 90000
Location: Paris, France
Starting date:
Urgent
Work mode: Hybrid
Published on: 25 March 2026
What they need
Job description
Verso is building the Consumer Intelligence OS.
We replace traditional qualitative research agencies with an AI-native platform that conducts, analyzes, and compounds consumer conversations at scale. Instead of one-off studies, companies get a continuous, structured understanding of their customers.
Our ambition is to make consumer insight a real-time, always-on system embedded in every product and marketing decision.
We're starting with AI-moderated interviews - but this is just the entry point. The long-term product is a system of record for consumer understanding, powered by agentic workflows.
Backed by Hexa (eFounders), behind Aircall, Front, Spendesk, Yousign and 50+ others.
Verso is founded by an experienced team with a strong track record in strategy, technology, and company building.
- CEO: HEC graduate, 7 years at BCG (Principal), expert in strategy, innovation, and consumer goods
- CTO: Dual degree X/HEC, former lead at CO2 AI, expert in AI and scalable product development
The role
You will join Verso as a Founding Engineer to own systems end-to-end from infrastructure to product, with no strict boundaries - and help define engineering standards, observability, and the foundations of the team.
You will design and scale a coherent AI system end-to-end, at the intersection of agents, data, and product:
- An AI interviewer interacting with users at scale (voice & video), sustaining long-form conversations with high reliability
- Agent systems (state, memory, evaluation loops) orchestrating complex interactions
- Real-time and async pipelines transforming multimodal data (video, audio, text) into structured, actionable insights
- A client platform to explore these insights through an AI-native interface
- Data systems to structure, query, and reuse qualitative knowledge at scale
- Optimization of latency, cost, and performance across LLM, TTS, STT
- Reliability of long-running AI workflows in production
What makes this role unique
- You will build production-grade AI systems, not prototypes
- You will work on long-form, high-stakes interactions, not short prompts
- You will design systems where AI quality is as critical as system reliability
- You will operate at the intersection of AI, data, infrastructure, and product
Current stack
- Frontend: TypeScript, React, Tailwind, shadcn
- Backend: Python, FastAPI, SQLAlchemy
- Infra: Vercel, Railway, Supabase AWS
- AI: LLM orchestration, STT/TTS pipelines, eval systems (rapidly evolving)
Profile wanted
Profile
We're flexible on background, but we're looking for people with a strong builder mindset.
- Some seniority (not straight out of school), but not necessarily 10+ years - 2-3 years in a strong environment is enough
- Experience in early-stage or high-ownership environments
- Strong bias toward action, autonomy, and ownership
We heavily overweight mindset over credentials:
- Product and user-first thinking
- Pragmatism and speed
- Optimism and resilience
- High intellectual honesty
- Strong drive
On hard skills, although Python and Typescript is a good +, but we don't optimize for specific languages. What matters is:
- Ability to design and scale production systems
- Comfort with ambiguity and messy real-world problems
- Ideally some exposure to AI / agentic systems in production (but not required)
TL;DR: we're looking for people who are sharp, driven, and with whom we could realistically build a company.
Preferred Experience
- Building production-grade AI systems on messy, real-world data (not demos)
- Owning systems end-to-end (architecture production iteration)
- Experience with complex systems (real-time, pipelines, or LLM-based products)
- Strong intuition for trade-offs: latency, cost, quality, reliability
- Motivation to build a category-defining company with long-term ownership
Why Join
- Category shift: we're turning consumer understanding into a continuous system - not just improving research workflows
- Hard problems: production AI over multimodal, unstructured data (audio, video, text, structured), across:
- a B2C product (low latency, high scale)
- a B2B platform (enterprise-grade reliability, security, integrity)
- Full ownership: architecture, product, and execution from day one
- Meaningful equity: early-stage package with strong upside (next fundraise soon)
- Hexa environment: access to top-tier talent, resources, and a strong builder ecosystem
Compétences requises
- Python
- Intelligence artificielle
- AWS
- TypeScript