In short
Lead AI Engineer in Sofia building AllVesta's Behavioural AI platform for financial institutions: designing production-grade agentic AI (multi-step reasoning, tool use, guardrails, evaluation frameworks), behavioural investor profiling, and scalable backend services using Python, AWS/Amazon Bedrock and GCP.
About the role
We're looking for the kind of engineer we'd want to start a company with.
Someone who thrives on complex challenges, takes ownership and has the ambition to create something that matters.
At AllVesta, we're on a mission to change how millions of people experience investing.
Our Behavioural AI platform helps financial institutions understand the human behind the money and turn savers into confident, long-term investors.
We're looking for an exceptional Lead AI Engineer to join us on that mission.
The ChallengeYou'll help shape and scale our Behavioural AI platform, working at the intersection of agentic AI, behavioural intelligence and financial services.
This is not about connecting an LLM to a chatbot. It's about engineering intelligent, reliable and explainable AI agents that can operate safely in regulated financial environments.
You'll have real technical ownership, from architecture and development to production deployment, working directly with leading financial institutions.
What You'll Do- Agentic AI Engineering: Design, build and evolve sophisticated AI agents, including multi-step reasoning, tool use, orchestration and adaptive conversational workflows.
- Agent Harness & Reliability: Engineer the systems behind our agents, including evaluation frameworks, guardrails, state management, observability, error handling and scenario replay.
- Behavioural Intelligence: Develop AI capabilities that support dynamic investor profiling, readiness assessment and personalised engagement.
- Platform Engineering: Build scalable backend services, APIs and enterprise integrations using Python and modern cloud technologies.
- AWS/GCP & Production Deployment: Architect, deploy and operate secure, production-grade AI systems using AWS/Amazon Bedrock and GCP.
- Technical Ownership: Take accountability for architecture, code quality, engineering decisions and reliable delivery from development through production.
- You've built AI agents and taken them into production, including the engineering harness around them: conversation state and memory, tool orchestration, and guardrails enforced outside the model.
- You understand which rules belong in code and which in prompts, and can explain the trade-offs.
- You know how to test agents rigorously, using evaluation datasets, automated judging and regression testing across prompt and model changes.
- You design for auditability, ensuring agent actions, decisions and supporting evidence can be traced and reviewed.
- You know how to optimise cost, latency and reliability as usage scales.
- You can set the technical direction of our AI platform, make architectural decisions and lead engineers as the team grows.
- You bring strong hands-on expertise in Python, AWS and Amazon Bedrock.
- You're experienced in backend engineering, APIs, cloud security and enterprise integrations.
- You combine independent thinking, sound technical judgment and a genuine sense of ownership.
Experience in financial services, healthcare or other regulated industries is a must.
Why Join AllVesta?Real ownership. Ambitious challenges. The opportunity to shape technology that matters.
You'll work alongside an experienced, founder-led team, helping advance an AI platform designed to reach millions of people through leading financial institutions.
You'll have the freedom to innovate, the responsibility to deliver and a direct influence on our technology and product direction.
This opportunity isn't for everyone. And that's fine.
But maybe you're not everyone.
Location: Sofia, Bulgaria
Apply: careers@allvesta.ai
- Tell us about the most ambitious AI system you've personally built and deployed into production. What did you own, how did you ensure its reliability, and what was the outcome?