AI Infrastructure Becomes the New Strategic Layer
AI infrastructure funding refers to investment in the foundational tools, databases, observability platforms, and safety systems that power AI agents and applications, sitting beneath end‑user products and enabling secure, scalable, and governed enterprise automation across industries such as software development, healthcare, and operations. This layer is now attracting serious capital and M&A attention. Across four recent deals, startups building AI coding infrastructure, AI observability tools, agentic AI databases, and responsible healthcare platforms have collectively secured USD 178 million (approx. RM819 million) in funding and acquisitions. The trend shows that investors and buyers are shifting from flashy AI demos to durable enterprise AI tooling. These companies are not consumer apps; they are the plumbing that lets AI agents work safely with code, systems, and patients. Their progress hints at a maturing market where infrastructure, not chatbots, becomes the primary battleground.
Niteshift and Elastic Show Enterprise AI Tooling Is Strategic
In software development and operations, AI infrastructure funding is clustering around tools that separate models from the platforms enterprises rely on. Niteshift, created by former early Datadog engineers, raised USD 7 million (approx. RM32 million) to build model‑agnostic AI coding infrastructure. Instead of replacing coding agents, Niteshift acts as a routing layer, orchestrating between frontier and open‑source models and charging per‑minute infrastructure fees, aligning closer to a cloud provider than a labour substitute. On the observability side, Elastic’s acquisition of DeductiveAI for up to USD 85 million (approx. RM391 million) shows how AI observability tools have become acquisition currency. DeductiveAI’s AI agents connect to code, logs, metrics, and traces, reasoning over a live knowledge graph to cut incident resolution times; DoorDash reportedly saved more than 1,000 engineering hours annually through this automation. Together, these moves show enterprise AI tooling is now core strategy, not a side experiment.
PhoenixAI Bets on the Agentic AI Database
For data infrastructure, PhoenixAI’s USD 80 million (approx. RM368 million) Series B underlines how central the agentic AI database has become. The company’s AI‑native engine is built for agentic AI workloads that fire thousands of unpredictable, real‑time queries, unifying live and historical data to deliver sub‑second responses at large scale. Customers such as AppLovin, Coinbase, Conductor, and Demandbase already run PhoenixAI in production, integrating it with Apache Iceberg data lakehouses and Kafka pipelines. As PhoenixAI’s leadership notes, today’s agents handle mission‑critical tasks, from customer service to supply chains, straining traditional analytics stacks that were never designed for this query pattern. By providing governance features for regulated industries, PhoenixAI positions itself as infrastructure that both data teams and compliance leaders can trust. In the landscape of enterprise AI tooling, it shows how agent‑ready databases are becoming as vital as classic data warehouses once were.
Vali Health Brings Responsible AI Infrastructure to Home Care
While much AI infrastructure targets engineering teams, Vali Health shows how responsible AI can reshape healthcare operations. The company emerged from stealth with USD 6 million (approx. RM28 million) to build AI infrastructure for home care agencies, a sector where coordination failures can have life‑and‑death consequences. Vali’s AI agents automate workforce management and complex scheduling, from handling no‑shows to matching specialised caregivers to clients with conditions like dementia or Parkinson’s. The platform runs a 24/7 safety infrastructure and is on track to complete 98% of tasks autonomously, with the remaining 2% routed to humans. Even fully autonomous tasks pass through a human‑in‑the‑loop review layer to maintain trust and safety. According to agency owner Dave Heinze, Vali is “the first thing that’s given my team time back,” saving mid‑sized agencies up to 20 hours per week and enabling growth without burning out staff.
A USD 178M Signal: Infrastructure Is Where AI Value Concentrates
Taken together, these four startups highlight how AI infrastructure funding and M&A is concentrating on the foundations that make AI practical in the enterprise. Niteshift targets AI coding infrastructure so enterprises can use diverse models without locking into model vendors that compete with them. Elastic’s acquisition of DeductiveAI shows that AI‑native ops tooling has become a must‑have feature in observability suites. PhoenixAI proves that an agentic AI database is now a distinct category, optimised for unpredictable agent traffic rather than human dashboards. Vali Health, meanwhile, shows that responsible AI infrastructure with strong guardrails can win trust in sensitive home care settings. Their combined USD 178 million (approx. RM819 million) in funding and exits sends a clear message: the next phase of AI will be defined by reliable, specialised infrastructure that enterprises can plug into core workflows, not by standalone chat interfaces.
