Position Overview
We are looking for a hands-on Backend Engineer to join the AI Center of Excellence at Wissen Infotech. You will own the design, development, and performance of the core Python services that power our AI Workbench—an internal platform that lets data scientists, prompt-engineers, and domain-experts rapidly experiment, train, evaluate, and deploy large-language-model (LLM) solutions at enterprise scale. Experience building AI / ML tooling is useful but not mandatory—we value rock-solid Python fundamentals, API-first thinking, and a passion for building developer-friendly products.
What You’ll Do
- Design and implement high-performance, async REST & WebSocket APIs in FastAPI that expose agentic, RAG, evaluation, and data-curation pipelines
- Build reusable service layers (authentication, authorization, rate-limiting, multi tenancy, audit) that accelerate new AI experiments while meeting enterprise-security mandates
- Integrate with vector DBs and LLM gateways
- Develop workflow orchestration jobs (Celery, Temporal, or Prefect) for long-running training, batch-evaluation, and data-labelling tasks
- Optimise throughput & latency (asyncio, connection-pooling, caching, gRPC) and protect downstream services with circuit-breakers & back-pressure
- Instrument code with Open-Telemetry, Prometheus metrics, and structured logs; create Grafana dashboards that help AI engineers spot regressions fast
- Partner with MLOps & DevOps teams to containerise services (Docker), define Helm charts, and drive CI / CD best-practices (GitHub Actions, ArgoCD)
- Write unit, integration, and property-based tests (pytest, hypothesis, locust) and maintain ≥80 % coverage in critical paths
- Publish clear API contracts (OpenAPI 3.x), SDK stubs, and “how-to” recipes in Confluence; mentor junior engineers via design reviews & pair-programming
Must-Have Skills (4-6 yrs)
Expert-level Python 3.9+ (type-hints, async / await, data-classes, pydantic)Production experience with FastAPI (dependency-injection, middleware, background tasks, SQLAlchemy-async)Solid RDBMS & NoSQL design (PostgreSQL, DynamoDB, Redis) including indexing, query-plan analysis, and migration tooling (Alembic)RESTful & real-time API design—status codes, idempotency keys, versioning, pagination, OAuth2 / JWT scopesTest-driven mindset : pytest fixtures, mocking, test-containers, CI gatesGit workflows (trunk-based or GitFlow) and peer-review disciplineLinux & container literacy—Dockerfile best-practices, multi-stage builds, docker compose for local dev parityBasic cloud fluency (AWS, Azure, or GCP) around IAM, KMS, S3 / Blob, Secret-Manager, Parameter-StoreCuriosity for AI concepts (prompts, embeddings, fine-tune jobs, RLHF) and willingness to upskill quickly Nice-to-Have (Preferred, Not Required)Hands-on with LLM frameworks (LangChain, LlamaIndex, Haystack) or inference servers (TGI, vLLM, Ray-Serve)Vector-database experience (Pinecone, Weaviate, Milvus, PGVector)GPU workload tuning (CUDA drivers, NCCL, batch-size optimisation)Event-driven architectures (Kafka, Pulsar, Redis-Streams)GraphQL (Strawberry, Ariadne) or gRPC service definitionsSecurity hardening : OWASP top-10, SAST / DAST, secrets-scanning, RBAC / ABACOpen-source contributions or technical bloggingQualifications
Bachelor’s degree in Computer Science, Software Engineering, or related field (or equivalent practical experience)4-5 years of progressive backend development in product-centric teamsDemonstrable track record of shipping scalable, well-tested Python services to production