The new additions—Tera Context Engine, Tera Harness and a suite of Agent Skills—are positioned as part of Teradata’s Autonomous Knowledge Platform and can be deployed together or as stand‑alone modules, according to the company.
Tera Context Engine is described as a context and orchestration layer that links databases, data platforms, pipeline engines, catalogs, models and AI agents without moving data, allowing a shared layer of business knowledge across both structured and unstructured sources while tying into governance, lineage and access controls.
The engine relies on a native context graph that maps metadata, semantics and business meaning as relationships, and it can read from and write back to existing systems of record, improving contextual accuracy over time. Industry‑specific knowledge models combine validated sector expertise with statistical AI models to reduce the need for agents to infer business meaning from scratch, a feature Teradata says is especially valuable in regulated sectors.
Tera Harness serves as the execution layer, maintaining context across workflows and coordinating tools, models, data and skills without manual routing. Built on a Go‑native engine and gRPC, internal tests showed the harness could support 512 concurrent agents on a single eight‑vCPU virtual machine and handle 279 tool calls per minute, while offering state management, checkpointing and human‑approval pauses for long‑running tasks.
In benchmark comparisons, Teradata reported that Tera used 73% fewer tokens, completed work 42% faster and cut total cost by 58% versus Claude Code on the SWE‑bench Pro test, and achieved a 53% lower cost per reliably solved task compared with Snowflake Cortex Code on the data‑eng‑bench benchmark. The platform also earned the highest Pass3 score on data‑eng‑bench and tied for top performance on ADE‑bench.
Agent Skills provide reusable functions for data engineering, analysis and data‑science tasks, callable by agents or directly via natural‑language prompts. The offering includes Platform Agents for workload tuning, compute sizing, telemetry and FinOps, as well as Analytics Agents for natural‑language query generation, SQL, Python and query optimisation.
Teradata will also offer AI Services to help customers identify use cases, configure industry knowledge models and move projects into production, aiming to reduce the experimentation phase. "Most enterprises are not starting from scratch with AI. They are dealing with tools that do not work together and a skills gap that makes those tools hard to use at scale," said Sumeet Arora, Teradata’s chief product officer.
The expanded Tera suite is scheduled to become generally available in the fourth quarter of 2026.