- Maryland’s TEDCO invested $250,000 in Syntes AI, an enterprise AI startup founded in 2024.
- Syntes AI, led by co-CEO Christopher Ramsey who previously built LATISTA, uses governed AI agents to execute workflows.
- Syntes claims significant accuracy and token reduction, but has not published third-party performance data from production deployments.
Maryland’s TEDCO has put $250,000 into Syntes AI, a 2024-founded startup whose pitch is that most enterprise AI deployments fail not from lack of model quality but from data fragmentation, the same siloed, disconnected systems that keep AI outputs from ever reaching production.
Co-CEO Christopher Ramsey previously built LATISTA, an enterprise data management platform, and he and co-CEO Andrei Lavrov are applying a similar logic here: unify the data layer first, then automate on top of it.
What Syntes draws against standard copilot tools is execution: the agents are designed to trigger actions in connected production systems, not just surface recommendations. Governance policies constrain what agents can do, and every action is logged for audit and rollback.
Syntes claims its approach delivers more accurate responses and reduces LLM token consumption compared with querying models directly.
Unifying fragmented data behind a single governed layer is a direct answer to the integration complexity and data debt that remain central friction points holding back enterprise AI pilots.
In January 2026, Syntes announced a native integration with OpenAI models, letting customers run OpenAI reasoning and generation against their own knowledge graph while keeping outputs within existing governance boundaries. The practical value for enterprise buyers is that the integration preserves explainability and human-in-the-loop controls at the point where OpenAI outputs touch production data, a requirement in regulated industries where model outputs need audit trails. Workflows the company targets include competitive pricing, inventory optimisation and fraud prevention.
TEDCO, as an instrumentality of the State of Maryland, runs seed funds ranging from $100,000 to $500,000, directed at early-stage technology companies with Maryland operations and a path to venture funding or self-sustaining revenue. The $250,000 Syntes investment sits at the midpoint of that range. To qualify, companies must be based in Maryland, have at least one full-time employee and have received no more than $15 million in prior funding, excluding TEDCO’s own capital.
The TEDCO check is small by venture standards, but it provides runway to extend the OpenAI integration and build toward a larger raise. Whether the live knowledge graph architecture holds up at the scale of a mid-market or large enterprise deployment, where data volumes and schema complexity multiply, is the question Syntes will need to answer before institutional investors move in. Syntes has not published third-party performance data from production deployments.
