AI Software Engineer, Agent Harness
EnCharge AI
IndiaremotePosted 12 days ago
Skill Required
Software-EngineerAI-EngineerMachine-Learning-EngineerBackend-EngineerPython-DeveloperAI-Harness-EngineerSenior-AI-Agent-EngineerAI-Agent-EngineerAI-Agents-EngineerAI-Agent-Systems-EngineerAI-Agent-Platform-EngineerApplied-AI-Agent-EngineerAI-Agentic-Systems-EngineerAI EngineerSoftware EngineerSoftware DeveloperAIsoftware engineeringObservabilityEngineeringsecuritybuildingQuery OptimizationPythonC++RustRAGandGenerative AIGoMachine LearningFulltime
Key highlights
- 10+ years of software engineering experience in backend systems or ML infrastructure
Role overview
EnCharge AI is developing a next-generation AI platform built on an in-memory-computing architecture that delivers a 10x improvement in compute energy efficiency and performance for AI inference workloads. The AI Software Engineer role focuses on owning the agent harness layer that connects agents to files, tools, and documents with proper permissions, memory, unattended execution, and real outputs. This involves designing and implementing core system components, maintaining safety and reliability, and ensuring seamless integration between different architectural layers.
Responsibilities
- Own the harness architecture end to end — agent loop, safe execution, context management, knowledge base, memory, permissions, orchestration, outputs, interfaces, observability — one component per layer, with clear interfaces so layers can be swapped
- Build the pieces with no open-source equivalent such as session semantics, enforced permissions, memory in a human-editable file, orchestrator, and outputs
- Keep pace with evolving models by adapting adapters, prompt formats, tool-call schemas, stop conditions, and conducting benchmarking and evaluation
- Make tool use reliable across models with varying tool-calling quality through validation, repair, retries, and fallback mechanisms
- Develop agents, tools, and MCP servers for internal and customer use cases, and review them for security before shipping
- Build the evaluation harness including task suites, regression runs on every model or harness change, and tracking of cost and latency per task alongside quality
- Define interfaces including session API, CLI, GUI, and endpoints that existing tools can connect to
Requirements
- 10+ years of software engineering experience in backend systems or ML infrastructure
- Strong Python and at least one systems language (e.g., Go, Rust, C++)
- Have shipped and operated an agent loop in production covering tool use, multi-step workflows, and unattended runs
- Hands-on experience with RAG, context management, and memory for LLM applications
- Experience with sandboxing, isolation, and permission models for automated systems
- Have run open-weight models personally and understand how quantization and serving choices affect model behavior
- Comfort in fast-moving, ambiguous environments with strong product instincts
Nice to have
- Contributions to open-source agent harnesses or coding agents
- Experience with agent benchmarks such as SWE-bench or Terminal-Bench and building internal task suites
- Familiarity with MoE serving concepts including expert placement, tensor parallelism, and quantization
- Expertise in observability for LLM systems
- Background in document parsing and indexing pipelines
- Desktop or GUI application development experience
Additional details
- Location: Bengaluru, Karnataka (or throughout India remote-friendly with travel)
- Company: EnCharge AI builds the next generation AI platform using an in-memory-computing architecture that improves compute energy efficiency and performance for AI inference workloads
- Original posted on Himalayas