AI Engineer

Tessera Labs•San Jose, CA

22h•$200,000 - $250,000

About The Position

Tessera Labs is developing an AI platform designed to transform how large enterprises operate. Unlike traditional approaches that involve lengthy and costly overhauls, Tessera acts as a transformation engine. It's a governed, multi-agent platform that understands an enterprise's processes, data, and code as a connected system, enabling changes in weeks rather than years. The platform is vendor-agnostic, working with systems like SAP, Salesforce, Workday, Oracle, Snowflake, and MuleSoft. Key challenges include ensuring governance, with every action being logged, traceable, and reversible due to customer regulatory requirements, and achieving generality, allowing the platform to function across diverse and complex enterprise landscapes it hasn't encountered before. Tessera Labs emphasizes that they sell a product, not a service, with their team focused on product success. The company has secured $60M in funding, led by Andreessen Horowitz.

Requirements

3+ years of experience building and operating production software.
Meaningful recent experience with Large Language Model (LLM)-powered systems.
Experience shipping agentic systems that real users depend on and managing incidents when they fail.
Fluency in current agent toolkits: tool calling, orchestration, context engineering, RAG, evals, tracing, and understanding their limitations.
Experience building retrieval systems against messy real-world corpora and measuring their performance.
Background in traditional ML, including supervised learning, feature engineering, and model evaluation, with the ability to apply it when beneficial.
Treating evaluation as an engineering discipline, not just a final report.
Thinking in terms of systems and customer outcomes, rather than solely model metrics.
Comfort with non-determinism and experience building reliably on top of it.
Strong Python and TypeScript skills.
Ability to ship quickly and verify thoroughly.
Interest in large, complex, and undocumented enterprise systems.

Nice To Haves

Hands-on experience with enterprise platforms like SAP, Salesforce, Workday, Oracle, Snowflake, MuleSoft, ServiceNow, including their APIs and extension models.
Experience with knowledge graphs, ontologies, or semantic models for complex real-world systems.
Experience with code analysis, program transformation, or automated refactoring at scale.
Experience with MCP, sub-agents, or agent skill/plugin architectures.
Background in distributed systems or workflow engines, particularly those handling partial failures.
Experience at an early-stage startup or as a founder, demonstrating adaptability to broad scope.
Familiarity with enterprise security and compliance realities (SSO, RBAC, segregation of duties, PII handling, SOC 2, data residency).

Responsibilities

Design and ship production agents for enterprise transformation work, including understanding landscapes, planning changes, executing them across process, data, and code, and verifying correctness.
Build the tool layer, creating typed, permissioned, and well-documented interfaces for agents to interact with enterprise systems safely.
Improve agent performance through techniques like prompting, context construction, tool-use strategy, and decision logic.
Develop the retrieval layer for enterprise artifacts and metadata, including chunking, indexing, hybrid search, reranking, grounding, and evaluation.
Manage context deliberately, deciding what information models see, compress, or drop, recognizing this as a critical engineering problem.
Utilize classical Machine Learning (ML) for tasks like routing, ranking, classification, anomaly detection, and confidence estimation where appropriate.
Build the evaluation and monitoring layer, including task sets from real customer landscapes, regression coverage, and alerting for system changes.
Instrument every run to ensure reconstructibility of model calls, tool invocations, decisions, and human approvals for auditability.
Diagnose production failures by analyzing execution traces and identifying root causes to improve the system.
Design guardrails, approval gates, and rollback paths to ensure agent safety when operating on live enterprise landscapes.
Build generalized capability that works across multiple customers, treating customer-specific solutions as product bugs.