Tekonika - Senior Full Stack AI Engineer

Job Description :

Role : Senior Full-Stack AI Engineer (AI Agent Platform)

Role Overview :

You will build and scale autonomous agent systems capable of querying data, executing sandbox code, managing marketing cohorts, and triggering cold outreach campaigns. Our architecture is a monorepo powered by Turborepo and Bun, running a LangGraph agent server and a Hono backend API, with a React 19 frontend and Kubernetes deployments. If you love building complex AI-native systems, optimizing data pipelines, and implementing high-fidelity interfaces, this role is for you.

Key Responsibilities :

- AI Agent Engineering : Design, implement, and optimize stateful multi-agent workflows using LangGraph and LangChain for marketing analytics, lead scoring, and automatic campaign generation.

- Backend API Development : Write fast, lightweight, and type-safe API services in TypeScript using the Hono framework on the Bun runtime.

- Data Integration & Warehousing : Query and manage marketing audiences and cohorts using Google BigQuery, and manage transactional application data in PostgreSQL via Prisma ORM.

- Real-time Frontend Development : Build reactive, accessible, and highly aesthetic user interfaces in React.js using TanStack Router, TanStack Query, and Radix UI/Base UI primitives.

- Asynchronous Processing : Build and scale background workers and job queues using Redis and BullMQ to handle high-throughput marketing data syncing and web scraping.

- Infrastructure & CI/CD : Manage containerized services using Docker and orchestrate them with Kubernetes (using Kustomize). Maintain automated CI/CD pipelines via GitHub Actions.

- Observability & Standards : Track LLM completions and cost performance with Langfuse, and maintain strict code quality standards using Ultracite (Oxlint + Oxfmt) and ESLint.

Technical Stack Requirements :

1. AI, LLM & Agent Orchestration (Core) :

- LangGraph & LangChain : 1. Extensive experience building stateful, cyclic graphs, memory savers (e.g., custom PostgreSQL checkpoints), and tool-calling schemas.

- Langfuse : 2. Experience with LLM tracing, evaluation, prompting registries, and analytics.

- APIs : 3. Integration experience with Anthropic (Claude Sonnet 4.6), Google Vertex AI (Gemini), and OpenAI APIs (e.g., Whisper, GPT-4o).

2. Backend & Database :

- Runtime : 1. Bun (as runtime, test runner, and package manager).

- Framework : 2. Hono (routing, middlewares, auth context).

- Database & ORMs : 3. Prisma ORM, PostgreSQL (Supabase / local), and Google BigQuery.

- Authentication : 4. Better Auth (handling organizations, session hooks, and service-to-service internal keys).

- Queues : 5. Redis & BullMQ (background job execution).

3. Frontend Client :

- Core : 1. React.js (React 19), Vite.

- Routing & State : 2. TanStack Router (file-based routing) and TanStack Query (server state synchronization).

- UI & Styling : 3. Radix UI / @base-ui/react primitives, Tailwind CSS (Tailwind v4 features), and Lucide Icons.

4. DevOps & Cloud :

- Docker & K8s : 1. Building multi-stage Dockerfiles and deploying manifests with Kubernetes Kustomize.

- Cloud : 2. Google Cloud Platform (GCP) and AWS S3 / Google Cloud Storage.

- CI/CD : 3. GitHub Actions (writing workflows, caching build layers, auto-formatting, and deploying to clusters).

Key Qualifications :

- 4+ years of professional experience in full-stack software development.

- 2+ years of experience working directly with LLMs, prompt engineering, and building agentic workflows (specifically LangGraph/LangChain).

- Deep understanding of TypeScript, monorepos, and modern build/runtime systems (like Turborepo and Bun).

- Experience handling database performance, writing raw SQL for BigQuery/PostgreSQL, and optimizing query speeds.

- Excellent understanding of CI/CD concepts and managing services on Kubernetes clusters.