Senior Engineer, AI Systems
Read straight from Tryzens Global's careers page, verified 3d ago.Seekless can tailor your résumé to it.
Seekless can write your résumé for this posting first.
Before you apply
LocationSofia, Sofia City Province, Bulgaria
Work moderemote
Typefull-time
SenioritySenior
Company size51–200 people
PostedJul 27, 2026 · 3mo ago
Verified live3d ago
At a glanceSummarised by Seekless from the posting.
Senior-level software engineering in Python/Typescript, with production ownership
Shipped an LLM or agent system that reached production, was used by people outside your team, and was maintained by you afterwards
Working method for knowing whether a change to a prompt, model or pipeline improved the system
Worked under data security or compliance constraints, including limits on what may be sent to third-party model providers
Understanding of failure modes specific to LLM systems - prompt injection, data leakage across tenants, unsafe tool invocation
Retrieval at scale: hybrid search, reranking, permissioned or fast-changing document sets
Agent frameworks, MCP
Bedrock / Vertex / Azure OpenAI
Asynchronous and long-running systems: queues, idempotency, durable state
Client-facing consulting
Skills
Python
Typescript
LLM
AI
MCP
Bedrock
Vertex
Azure OpenAI
REST APIs
Benefits
Health insurance
Flexible hours
Remote work
Vacation days
Meal vouchers
About the company
Tryzens Global is a digital commerce agency that helps the world’s most ambitious retailers scale their businesses internationally. We are the only global support partner in digital commerce operating across every time zone, with deep experience solving complex retail integrations around the world.
We provide best of breed solutions and services to global businesses within retail, wine & spirits, beauty & cosmetics, automotive, financial services and other industries. We have built our delivery focused reputation upon technical innovation, in-depth business knowledge, and creative vision, all of which supports our objective of helping clients to gain true value from their technology stacks. We have a platform-neutral independent approach working with the world’s leading technology partners like Shopify, Agentforce(Salesforce), SAP & Adobe.
About the role
We are embedding artificial intelligence across every stage of our delivery lifecycle and are looking to grow our AI Practice in Sofia. We are looking for an experienced , smart and energetic person to join our team of talented professionals as a Senior AI Developer.
The AI Developer role is a hands-on engineering role within the Tryzens AI Practice and is focused on building AI-augmented delivery capabilities across our core technology platforms.
Responsibilities
● Design, build and operate LLM- and agent-based features in production - architecture through deployment, monitoring and iteration - with cost and latency treated as design constraints, not later optimisation
● Integrate models with our systems and data: APIs, tool calling, multi-step orchestration, retrieval where the task requires grounding in specific data
● Build the evaluation and observability that make behaviour changes visible before release rather than after: test sets, automated scoring, tracing, per-task cost
● Design for non-deterministic failure - output validation, fallbacks, degradation paths, human checkpoints where an error cannot be absorbed
● Work with clients and product owners to turn ambiguous requirements into scoped, testable capabilities, including judging when an LLM is the wrong tool
Requirements
● Senior-level software engineering in Python/Typescript, with production ownership: you have deployed, monitored and been accountable for systems in live use
● You have shipped an LLM or agent system that reached production, was used by people outside your team, and was maintained by you afterwards
● You have a working method for knowing whether a change to a prompt, model or pipeline improved the system
● You have worked under data security or compliance constraints, including limits on what may be sent to third-party model providers, and you understand the failure modes specific to LLM systems - prompt injection, data leakage across tenants, unsafe tool invocation
Useful, not required
● Retrieval at scale: hybrid search, reranking, permissioned or fast-changing document sets
● Agent frameworks, MCP
● Bedrock / Vertex / Azure OpenAI
● Asynchronous and long-running systems: queues, idempotency, durable state
● Client-facing consulting
Not required
An ML or research background. This is an engineering role: the work is building systems around models, not developing them. No specific commerce platform experience.
Benefits
● Competitive remuneration package
● Flexible working hours
● Fully remote working policy
● Food and gift vouchers
● Additional health insurance
● Great work-life balance and 25 days annual vacation
● 1 day Birthday vacation
● A positive and fun environment