Product Engineer, Applied AI

As a Product Engineer in Applied AI at Pocus, you will be responsible for architecting and implementing intelligent agents that extract insights from diverse data sources, enhancing workflows that significantly support sales operations. You'll engage in building a flexible data ingestion layer, a semantic enrichment engine leveraging large language models, and a query engine for low-latency data processing. Your role is crucial in ensuring the security and reliability of systems while elevating engineering standards through clean abstractions and effective testing.

Ideal candidates for the Product Engineer, Applied AI position at Pocus should have at least 3 years of experience in building production-ready distributed systems, alongside 1 year of experience in production-grade AI applications. You should be adept at working with data-rich systems and comfortable reasoning across messy datasets. Additionally, familiarity with large language models, intelligent agent workflows, and various productivity tools is essential to thrive in this role.

At Pocus, the tech stack for the Product Engineer, Applied AI position includes AWS infrastructure, with services designed in TypeScript using GraphQL for seamless data integration. You'll also work with tools such as Clickhouse, Presto, and Temporal, alongside large language models as core components, giving you ample opportunity to experiment and grow your technical expertise.

The work culture at Pocus is vibrant and collaborative, defined by a community of humble overachievers who thrive in an environment that balances autonomy, high impact, and fun. We call it 'shiperate’—where quick iteration and ownership of projects are welcomed. You'll find a strong emphasis on mentorship, professional growth, and a commitment to maintaining a healthy work-life balance.

At Pocus, as a Product Engineer in Applied AI, you'll have multiple avenues for career growth. Our mentorship programs, supported by First Round Capital, offer personal growth opportunities. You'll gain hands-on experience in an emerging field, allowing for advancements in your technical skills and leadership abilities, paving the way for future career advancements within the company or the broader AI ecosystem.

To build an intelligent agent for sales workflows, I would start by understanding the specific needs of the sales team and the data they frequently utilize. Next, I would design a system that integrates key data sources, ensuring seamless data ingestion and enrichment. I would also focus on implementing machine learning models that assist in analyzing customer behaviors and predicting next steps, ultimately streamlining the sales process.

My experience with large language models includes experimenting with various applications, such as chatbots and automation for data analysis. I’ve leveraged LLMs to provide context-aware outputs and drive insightful decision-making processes. It has been rewarding to observe how they can simplify complex dataset interactions for end-users.

Certainly! One challenging project involved designing a distributed system for real-time data processing. It required managing data consistency and addressing latency issues. I focused on optimizing resource allocation and ensuring fault tolerance across nodes, which ultimately led to improved performance metrics and user satisfaction.

To ensure the reliability and security of my AI systems, I prioritize best practices such as thorough testing, including unit tests and integration tests. Additionally, I implement security measures like data encryption and access controls to safeguard sensitive information. Regular security audits and adherence to industry standards further enhance the robustness of the systems.

When juggling multiple projects, I prioritize tasks based on factors such as project deadlines, impact on customer outcomes, and required resources. I often use Agile methodology to maintain flexibility, hold regular check-ins with team members, and adjust priorities as necessary to ensure deliverables are met effectively.

I regularly use tools like GitHub Copilot for code suggestions and task automations, along with Jira for project management and collaboration. Code review platforms are also essential for maintaining code quality. These tools help me stay organized, streamline my workflow, and focus on more complex problem-solving tasks.

Feedback is an invaluable part of the development process. I approach it with an open mind and an understanding that it can lead to better outcomes. I typically engage in discussions with my peers to dissect the feedback, analyze it critically, and incorporate constructive suggestions into my work to align with project goals.

I have extensive experience with data ingestion and transformation, particularly in building pipelines that extract data from various sources. I've employed ETL (Extract, Transform, Load) processes to clean and prepare data for analysis, ensuring that it is structured appropriately for downstream applications, such as model training and reporting.

Collaboration is crucial in AI projects due to the multidisciplinary nature of the work. It fosters the exchange of ideas, innovation, and problem-solving, enabling teams to combine different perspectives and skills. Frequent communication ensures alignment of goals and facilitates knowledge sharing, which ultimately leads to better project outcomes.

I thrive in fast-paced environments because they challenge me and promote my personal growth. The dynamic nature of such an environment pushes me to think creatively and adapt quickly. Additionally, I find motivation in working as part of a team where every member is equally committed to achieving our shared goals, as seen at Pocus.