Software Engineer — Agentic Search Systems

The dawn of Large Language Models and autonomous AI agents has completely transformed how software engineering, information retrieval, and search systems operate. Traditional keyword matching and static index search are rapidly giving way to agentic workflows where AI models dynamically query, reason, and synthesize information across vast data stores. If you are an experienced software engineer who has designed, shipped, or optimized production search architectures in this new era, an elite remote opportunity is calling your name.

Mercor is actively recruiting top-tier technical minds for the Software Engineer — Agentic Search Systems contract position. Offering compensation ranging from $80 to $150 per task—supplemented by unique referral bonuses and paid evaluation interviews—this role provides unprecedented flexibility and compensation for engineers who understand the mechanics of modern retrieval systems.

Deep Dive: What Are Agentic Search Systems and Why Do They Matter?

For decades, enterprise search engines relied on structured databases, inverted indexes, lexical scoring (like BM25), and basic vector embeddings. While these systems served web and enterprise apps well, they struggle with multi-step reasoning, dynamic tool usage, and iterative information synthesis.

Enter agentic search. In an agentic architecture, the search system is no longer a static query processor; it is an autonomous loop where an LLM agent breaks down user queries, generates sub-queries, executes searches across multiple APIs or document stores, filters results, evaluates data relevance, and synthesizes an answer. Building and scaling these architectures requires a rare blend of classical retrieval engineering and cutting-edge LLM orchestration.

As a Software Engineer specializing in Agentic Search Systems with Mercor, your insights, architectural decisions, and production experiences will directly influence how next-generation AI models interact with data.

Core Technical Competencies in Demand:

• Production Retrieval Ownership: Demonstrable history of owning relevance ranking, latency optimization, and recall metrics on systems relied upon by real users at scale.
• Evaluation & Impact Metrics: Proven ability to quantify system improvements, design robust evaluation benchmarks, and measure retrieval quality objectively.
• Agentic Workflows: Hands-on experience building, debugging, and refining multi-step LLM search agents, tool-calling pipelines, and autonomous query expansion frameworks.
• Data Infrastructure Scaling: Expertise in engineering high-performance data pipelines, vector databases, and caching layers to power lightning-fast modern search experiences.

The Mercor Interview Experience: What to Expect

Traditional tech hiring is notoriously broken—requiring weeks of take-home projects, multiple rounds of whiteboard coding, and endless committee approvals. Mercor has streamlined this entire paradigm into a fast, transparent, and developer-friendly process:

1. Application Submission: Submit your professional profile detailing your experience with search systems, data infrastructure, and AI agents.
2. The 25-Minute Conversational Interview: No coding tests or take-home assignments. Instead, you engage in a streamlined conversation discussing how you think about search quality, evaluation, and system bottlenecks. Bring your specific architectural challenges and "war stories."
3. Paid Live Follow-Up ($200 Bonus): If your initial interview stands out, Mercor invites you to a paid 30-minute live conversation with their engineering team, compensating you $200 for your time upon call completion.

Compensation Structure and Earning Potential

Freelance software engineers and technical consultants thrive on transparency and high return on time invested. Mercor’s compensation model for this role is structured to reward high-caliber expertise:

• Task-Based Earnings: Earn between $80 and $150 per completed task, allowing you to scale your income dynamically based on output and availability.
• Paid Interview Milestones: Get paid for your expertise even before taking on project tasks, thanks to paid follow-up discussion milestones ($200).
• Lucrative Referral Program: Earn up to $600 for each successful professional referral through your unique link, with no caps on total referral earnings.

Whether you are supplementing a full-time engineering role, consulting independently, or building your own portfolio of technical projects, Mercor provides an agile income stream backed by top-tier AI labs.

Step-by-Step Guide to Completing Your Application Successfully

Because positions at the cutting edge of AI agent engineering attract intense global competition, your application must immediately convey your depth of experience. Follow these steps to maximize your acceptance odds:

• Highlight Production Systems: Focus your resume and application summary on real systems you've built or scaled. Mention specific metrics—such as latency reduction, QPS capacity, or relevance score improvements.
• Emphasize Search Architecture: Clearly articulate your familiarity with embedding models, vector databases, reranking algorithms, and LLM orchestration frameworks (e.g., LangChain, LlamaIndex, custom agent loops).
• Be Ready to Talk Tradeoffs: Prepare brief, articulate summaries of architectural challenges you've faced (e.g., how you handled stale index updates or mitigated hallucination loops in agentic retrieval).

Tech roles in agentic search are filling rapidly as AI labs race to deploy smarter retrieval systems. Don't leave your career progression to chance—submit your profile today.