Senior Agentic AI Engineer at Skylabs AI

Islamabad, Islamabad Capital Territory, Pakistan -

Full Time

Start Date

Immediate

Expiry Date

01 Apr, 26

Salary

0.0

Posted On

01 Jan, 26

Experience

5 year(s) or above

Remote Job

Yes

Telecommute

Yes

Sponsor Visa

Skills

Agentic AI, C++, Python, LLMs, Telemetry, Static Analysis, Dynamic Analysis, Docker, Cloud Infrastructure, API Design, Neurosymbolic AI, Software Engineering, Data Pipelines, Remote Execution, Build Systems, Debuggers

Industry

Research Services

Description

About Us At SkyLabs AI Inc., we are at the forefront of the artificial intelligence revolution. As a US-headquartered company, we conduct applied research on AI for intelligent reasoning. We specialize in complex neurosymbolic AI to solve intricate problems within software engineering. Our team is composed of world-class researchers and engineers dedicated to building the platforms and intelligent agents that will power the next generation of software. If you are passionate about building truly intelligent systems and want to make a lasting impact, join us. The Role We are seeking an exceptional Senior Agentic AI Engineer to join our core team. This is not a typical AI role; you will be responsible for architecting, building, and deploying complex, autonomous AI agents designed to reason about and operate within the C++ software engineering lifecycle. You will be building agents that can programmatically interact with compilers (LLVM), debuggers (GDB/LLDB), build systems (CMake/Bazel), and static analysis tools to understand, modify, and build complex C++ codebases. The ideal candidate is a master programmer with a deep, practical understanding of both agentic AI and the C++ developer ecosystem. Key Responsibilities Architect & Build Agents: Design, develop, and refine sophisticated, multi-step AI agents that can reason, plan, and execute complex software engineering tasks. Integrate SWE Tools: Natively integrate agents with the C++ ecosystem (build systems, debuggers, profilers, and static analyzers) to enable programmatic code comprehension and modification. Develop Agentic Frameworks: Implement and innovate on agentic patterns (ReAct, CoT, LangGraph) and orchestrate multi-agent systems. Build Data Pipelines: Design telemetry systems to capture agent trajectories (state, action, observation) to create high-quality datasets for fine-tuning. Manage RAG Systems: Implement and optimize RAG pipelines using vector databases and code-specific embedding models (e.g., GraphCodeBert) for knowledge retrieval. Design Sandboxed Environments: Develop remote execution environments (using Docker, Firecracker) where agents can safely build and run C++ code. Collaborate on Infrastructure: Work with the backend team to design APIs (gRPC, REST) and cloud infrastructure (Kubernetes, GCP/AWS) to support the agentic systems. Qualifications & SkillsCore Profile: Agentic AI Developer Excellent theoretical and practical knowledge of LLMs: Deep understanding of transformer architectures, attention mechanisms, fine-tuning techniques (LoRA, QLoRA), and retrieval-augmented generation (RAG). Vast experience in building LLM agents: Proven track record of designing, implementing, and evaluating complex, multi-step agents that can reason, plan, and use tools. Excellent development skills in Python: Mastery of Python for AI/ML development, data processing, and backend service integration. Strong Computer Science (CS) background: Solid foundation in algorithms, data structures, operating systems, and software engineering principles. Specific Experience Required: 1. Agentic Frameworks & LLM Operations Advanced Prompt Engineering: Demonstrable experience with reasoning-and-action (ReAct), Chain of Thought (CoT), and graph-based agent structures (e.g., LangGraph) to handle complex, long-horizon tasks. Agent Orchestration Frameworks: Hands-on experience with frameworks like LangGraph for coordinating multiple agents and managing complex state. RAG: Vector databases, embedding models for code (e.g., GraphCodeBert) and Graph Vector DBs based on ASTs. LLM Evaluation & Benchmarking: Experience with code generation benchmarks, specifically SWE-bench (and knowledge of its variants like SWE-bench++), to rigorously test and improve model performance. Telemetry & Fine-Tuning Pipelines: Designing data collection systems (telemetry) to capture agent trajectories (state, action, observation) for creating high-quality datasets for model fine-tuning. 2. C++ & Software Engineering (SWE) Tooling Deep C++ Ecosystem Knowledge: This is non-negotiable. Must have experience with the tools the agent will use, including: Build Systems: CMake, Make, and preferably Bazel. Debuggers: GDB, LLDB, and their programmatic interfaces. Profilers: perf, Valgrind, or other C++ performance analysis tools. IDE Extension / Language Server Protocol (LSP): Experience building IDE plugins (for VS Code, CLion, etc.) or familiarity with the Language Server Protocol (LSP) to understand how an agent can programmatically interact with an IDE's core features (e.g., "find all references," "go to definition"). 3. Program Analysis Static & Dynamic Analysis: Understanding of static analysis (e.g., Clang-Tidy, Clang Static Analyzer) and how to parse their output to guide an agent's code modifications. Abstract Syntax Trees (ASTs): Experience using tools like Clang's LibTooling or Tree-sitter to programmatically parse, analyze, and perform "surgical changes" to C++ code, which is far more reliable than text-based regex. 4. Backend & Cloud Infrastructure Remote Execution & Sandboxing: Experience with container technologies (Docker, Firecracker) and remote execution environments to create the "remote sandbox" where agents can safely build and run code. API & Protocol Design: Experience designing and integrating with RPC/API protocols (gRPC, REST) to handle communication between the IDE plugin, the backend, and various agent servers. Cloud Services & Databases: Proficiency with a major cloud provider (AWS, GCP, Azure), database management (SQL/NoSQL), and cloud-native orchestration tools (e.g., Kubernetes) for managing the agent orchestration and backend infrastructure. Who You Are You are a builder at heart, driven by curiosity and a desire to solve hard problems. You don't wait for instructions; you identify opportunities and seize them. You are comfortable with ambiguity and thrive in a fast-paced, innovative environment where you are encouraged to experiment and push boundaries. You are a clear communicator who can articulate complex technical concepts and work collaboratively to achieve ambitious goals. What We Offer Competitive salary in USD Comprehensive health allowance Relocation allowance (if you're moving to Islamabad) Monthly team events and offsites A beautiful, collaborative office space Work alongside world-class AI researchers and engineers

Responsibilities

The Senior Agentic AI Engineer will architect, build, and deploy complex autonomous AI agents for the C++ software engineering lifecycle. Responsibilities include integrating with software engineering tools and developing agentic frameworks.