Senior Software Engineer, Agent

Senior Software Engineer, Agent

Department: Engineering | Location: Palo Alto HQ | Type: Full-time, On-site

About the Role:

We're looking for a Senior Agent Engineer to push the boundaries of what AI agents can do at Pika. You'll work on the systems that give AI agents their personality, memory, reasoning, multi-modal capabilities, and ability to act autonomously across platforms. This is the core of what makes Pika's AI products feel alive.

You'll design agent architectures, build tool-use frameworks, optimize LLM interactions, and create the systems that allow agents to learn, remember, and evolve over time. If you're excited about building AI systems that feel genuinely intelligent — not just wrappers over chat APIs — this role is for you.

• What You'll Do

• Design and evolve the agent runtime — the core loop that handles reasoning, tool use, memory retrieval, and response generation

• Build agent capabilities — image generation, voice synthesis, video creation, web browsing, code execution, and other skills

• Optimize LLM integration — prompt engineering, context window management, multi-provider model routing (Claude, GPT, Gemini, open-source), cost optimization, and latency reduction

• Implement memory systems — long-term memory, working memory, episodic recall, and semantic search so agents learn from every interaction

• Design autonomous behaviors — proactive actions, scheduled tasks, and goal-directed behavior that makes agents feel self-directed

• Build evaluation and quality systems — benchmarks, A/B testing, and metrics for agent behavior, personality consistency, and response quality

• Experiment with new architectures — multi-agent collaboration, planning, chain-of-thought reasoning, and other emerging patterns

• Collaborate with product and design to translate AI capabilities into intuitive user experiences

• What We're Looking For

• 5+ years of software engineering experience, with 2+ years working with LLMs or AI agent systems

• Deep understanding of LLM capabilities and limitations — you know how to get the best out of frontier models

• Experience building agent systems — tool use, function calling, multi-step reasoning, retrieval-augmented generation (RAG)

• Strong prompt engineering skills — system prompts, few-shot learning, chain-of-thought, structured output

• Proficiency in TypeScript and/or Python

• Understanding of embedding models and vector search for memory and retrieval

• Comfortable with rapid experimentation — you ship experiments, measure results, and iterate

• Product intuition — you understand what makes an AI agent feel "alive" vs. robotic

• Clear communication skills and a team-first mindset

• Nice to Have

• Experience with multi-modal AI (image generation, TTS, speech-to-text, video generation)

• Experience with agent frameworks (Lang Chain, AutoGPT, CrewAI, or custom runtimes)

• Background in NLP, computational linguistics, or cognitive science

• Experience with fine-tuning, RLHF, or DPO

• Familiarity with AI safety and alignment considerations

• Experience with real-time/streaming LLM responses

• Previous startup experience — comfortable with ambiguity and moving fast