Lead Full Stack Developer (AI Native, Python + TypeScript)

Job description

About the job

We are seeking an AI-native Full Stack Lead with strong expertise across the modern Python stack (Django, FastAPI) and the modern frontend stack (Angular, React, TypeScript) to lead the design, development, and optimization of end-to-end systems — systems where AI is not a side feature but a core capability of the product.

This is a modern engineering leadership role with two intertwined sides.

First, the lead uses AI to build: hands-on with agentic engineering workflows — Claude Code and similar tools — while setting the standards, harnesses, and guardrails that allow AI agents to safely contribute code at scale.

Second, the lead builds AI into the product: integrating LLMs, image models, embeddings, RAG pipelines, and agentic features directly into customer-facing experiences.

The ideal candidate guides the team in shipping scalable, secure, and high-performance APIs, services, user interfaces, and AI capabilities; mentors engineers on how to work effectively alongside coding agents and how to embed AI responsibly into products; and drives technical excellence in a world where humans set the spec and govern the work, and agents do a large share of the building.

Key Responsibilities

• Lead the architecture, design, and development of end-to-end systems — backend services in Python (Django and FastAPI) and frontend applications in Angular, React, and TypeScript.

• Lead the design and delivery of embedded AI capabilities in the product — LLM-powered features, RAG-backed assistants, image generation/understanding, embeddings-driven search and recommendations, and in-product agents.

• Make architectural calls on model selection, hosting, and orchestration: which LLMs, image models, and embedding models to use, when to use proprietary APIs vs.

open-source, and how to balance quality, cost, and latency.

• Build and maintain scalable RESTful APIs with high performance and security standards.

• Set the bar for frontend architecture, component design, type safety, and UX quality across Angular and React codebases — including the UX patterns specific to AI features (streaming, citations, partial outputs, error and refusal states).

• Oversee integration with third-party services, AI APIs (Anthropic, OpenAI, Google, open-source model providers), and internal tools.

• Ensure best practices for security, authentication, data protection, and responsible AI across the stack.

• Optimize system performance and scalability — backend, frontend, AI inference paths, and the seams between them.

• Manage cloud deployments (AWS, Azure, or GCP) and containerization (Docker, Kubernetes).

• Collaborate with product, design, and engineering teams to deliver features end to end.

• Mentor and coach full-stack developers, conduct code reviews, and define technical guidelines — including patterns for embedding AI safely into product surfaces.

• Drive adoption of CI/CD, automated testing, evals for AI features, and software engineering best practices.

• Lead the team’s adoption of agentic engineering — establishing how Claude Code and other coding agents are used to design, implement, and review systems across the stack.

• Define and maintain the harnesses, context files, and guardrails (spec documents, repo conventions, test gates, review policies) that govern AI-agent contributions to the codebase.

• Champion spec-driven development using BMAD, GSD, or similar open-source frameworks so that agents and humans build against the same source of truth.

• Continuously evaluate new agentic tools, models, and workflows, and decide what the team adopts, adapts, or rejects — both for how we build and what we build.

Requirements

• Bachelor’s or Master’s degree in Computer Science or related field.

• 4+ years of full-stack development experience, with 2+ years in a leadership role.

• Strong expertise in Python with both Django and FastAPI.

• Strong frontend expertise with TypeScript and at least one of Angular or React — ideally both, or willingness to operate fluently across both.

• Hands-on experience with databases (PostgreSQL, MySQL, MongoDB).

• Solid understanding of microservices, APIs, and distributed systems.

• Strong cloud deployment experience (AWS/Azure/GCP).

• Track record of shipping at least one production product with embedded AI capabilities (LLM features, RAG, agents, image, or similar).

• Excellent communication, leadership, and problem-solving skills.

Building AI-Native Products

These roles ship products that have AI inside.

We expect the lead to set the bar for the team across the full stack of AI capabilities a modern product depends on:

• Hands-on experience integrating AI APIs (Anthropic Claude, OpenAI, Google Gemini, and open-source providers) into production backends — including streaming, function/tool calling, structured outputs, and retries.

• Working literacy across model families: large language models, multimodal/vision models, image generation models (e.g. Stable Diffusion, Flux, DALL·E), speech models, and embedding models.

Knows when each is the right tool, and what each costs in quality, latency, and dollars.

• Foundational understanding of how modern ML works: tokens, context windows, embeddings, fine-tuning vs.

prompting vs.

RAG, and the trade-offs between them.

Doesn’t treat models as black boxes.

• Deep, hands-on experience designing and shipping RAG systems: chunking strategies, embedding choice, vector databases (pgvector, Pinecone, Weaviate, Qdrant, or similar), hybrid retrieval, reranking, and retrieval evaluation.

• Experience designing and shipping in-product agentic features — features where the AI plans, calls tools, and acts on the user’s behalf rather than just answering.

• Hands-on experience building multi-step agent flows — planning, tool use, memory, and reasoning — using frameworks such as LangChain, LangGraph, CrewAI, AutoGen, or equivalent.

• Production-grade prompt engineering: system prompts, structured outputs, ReAct / chain-of-thought / reflection patterns, and prompt versioning.

• Strong grasp of context engineering: building and curating the context windows, retrieval pipelines, and memory stores that make AI features reliable rather than impressive-once.

• Working knowledge of the Model Context Protocol (MCP) — consuming MCP servers, and ideally building custom ones to expose internal tools and data to agents safely.

• Ability to design and run evals for AI features — golden datasets, LLM-as-judge, regression suites, and CI gates that catch quality drift before users do.

• Experience with AI observability and tracing tools (LangSmith, Langfuse, Braintrust, Phoenix, or similar) to debug multi-step runs and monitor production AI features.

• Practical understanding of guardrails — pre-LLM input filtering, post-LLM output validation, PII/secret redaction, schema enforcement, and safe-action policies.

• Experience designing AI UX patterns: streaming, citations, confidence cues, partial states, error and refusal handling, and human-in-the-loop checkpoints.

• Awareness of the security and governance surface area of AI products: prompt injection, tool-use abuse, data exfiltration risks, hallucination management, and how to mitigate them.

• Comfort with cost, latency, and reliability trade-offs across model choice, context size, caching, and orchestration patterns.

Agentic Engineering Qualifications (Using AI to Build)

Beyond building AI into the product, we expect the lead to be fluent in using AI to build the product:

• Hands-on experience building real applications with Claude Code and/or other coding agents (Cursor, Aider, Cline, Devin, or similar) — not just experiments.

• Working experience with spec-driven development using BMAD, GSD, or other open-source agentic frameworks.

• Demonstrated ability to keep up with the rapidly changing AI-engineering landscape — reading, experimenting, and forming a point of view on what works.

• Openness to try new tools and workflows based on what the team is learning, and willingness to discard approaches that no longer hold up.

• Acceptance that AI agents will be writing a significant share of the code, and a clear sense that the human’s job is to govern that work — designing harnesses, curating context, setting guardrails, and reviewing outputs against the spec.

Nice to Have

• Experience fine-tuning models or working with model adapters (LoRA, PEFT, or similar).

• Experience with analytics tools and monitoring systems.

• Experience with mobile or cross-platform frameworks (React Native, Flutter, or similar).

• Experience defining evals, golden tests, or other automated checks for AI-generated code.

• Contributions to open-source agentic tooling, MCP servers, or spec-driven frameworks (BMAD, GSD, or similar).

• Recognised certifications such as IBM RAG and Agentic AI Professional, NVIDIA Agentic AI / LLMs, or equivalent — useful but not a substitute for shipped work.
  
  

Required Skill Profession

Social Networking Platforms