How The Intelligent Search Company is training AI to act with split-second intuition

The Intelligent Search Company, also known as TISC, has formally emerged from stealth with a transformer-based AI platform designed to power real-time decision-making in high-stakes environments. The Bay Area and Toronto-headquartered startup also disclosed that it raised US$2.1 million in pre-seed funding led by OVO Fund and n49p, with participation from Panache Ventures. Founded in May 2024 by Arpan Bhattacharya and Mahbod (Moe) Sabbaghi, TISC aims to give human teams and AI agents “split-second intuition” by fusing search, reasoning, and prediction into one real-time engine.

TISC’s technology is specifically built to parse dynamic, fragmented data and deliver relevant insights without delay. Target sectors include sports, emergency services, and defense, which are areas where milliseconds can alter outcomes. Unlike traditional large language model tools that rely on keyword matching or slow, multi-query logic chains, TISC’s system claims to surface precise, actionable context by understanding both the user’s intent and the situation’s urgency.

This real-time retrieval platform was developed after the founders recognized a key gap in AI readiness: existing agents were too slow, brittle, or verbose to support high-pressure decisions. With generative AI tools now widely deployed in productivity settings, TISC is positioning its transformer architecture as a foundation for the next generation of action-oriented AI—ones that can support coaches, commanders, or medics under real-world constraints.

Why real-time decision intelligence needs a different kind of AI retrieval model

The core problem TISC tackles is the inefficiency of standard AI search in situations where delays are unacceptable. Most AI agents today operate by issuing one or more queries, retrieving document-level data, and then reasoning over that information. This traditional separation between retrieval and reasoning slows the process down, introduces uncertainty, and makes the system ill-suited for mission-critical applications.

TISC’s model takes a fundamentally different approach. Instead of stacking a search engine on top of a generative model, the startup trains a transformer to learn what data is relevant, when to recall it, and how to act on it. The system continuously scans fast-moving data across formats such as video, telemetry, sensor feeds, or scouting reports and synthesizes insights in context. This allows the AI to predict opponent actions in sports, suggest countermeasures in defense scenarios, or identify optimal routes for emergency responders without requiring human operators to refine their queries or navigate complex dashboards.

This integration of real-time memory and context prediction allows TISC to function more like a human assistant with domain fluency than a generic AI chatbot. The architecture supports situational awareness and counter-decisioning, effectively enabling what the founders call “AI intuition.” The goal is to close the latency gap between information availability and decisive action.

Inside TISC’s early sports pilots and why the Women’s Premier Basketball Association is backing the technology

TISC has already tested its technology in live environments, beginning with pilots in collegiate and national sports teams. Notably, the platform was deployed during NCAA Division 1 games and used by the national basketball team of a Caribbean island nation. The most visible partnership so far is with the Women’s Premier Basketball Association, or WPBA, which is now preparing to integrate TISC across multiple touchpoints within its training, scouting, and gameplay environments.

WPBA, a developmental league focused on preparing elite women athletes for international play and the WNBA, sees TISC as central to its AI-first vision. By 2026, the league plans to introduce a real-time assistant coach powered by TISC that can answer spoken English queries during games and practices. These queries will surface tactical video clips, statistics, and adjustment recommendations in seconds without the need for filtering through spreadsheets, tagging systems, or multiple software platforms.

According to WPBA founder and chief executive Faatimah A., the goal is to create a tech-enabled environment where player development is informed by precise, accessible data. With TISC, coaches can make real-time substitutions or play-calling changes based on data gathered from scouting reports, match footage, or evolving game dynamics. The platform is also being used to help WPBA scouts centralize film reviews, live performance indicators, and developmental notes into a single searchable interface. The intention is to make the league’s roster more accessible to WNBA recruiters and remove friction in the talent evaluation process.

What sets TISC apart from existing LLM-powered tools and retrieval-augmented generation platforms

In a market saturated with large language model integrations and retrieval-augmented generation frameworks, TISC is betting on deeper architectural innovation. Most generative platforms today rely on a retrieval step that pulls in documents and a generation step that summarizes or responds. This step-wise approach works in general-purpose search but is inadequate for time-sensitive operations where users cannot afford trial-and-error prompts or irrelevant results.

TISC eliminates this separation by embedding memory and prediction directly into the transformer. It learns how to associate live data inputs with desired outcomes, reducing the need for blind follow-up queries and optimizing for latency. This is particularly impactful in multi-modal contexts, where input sources include not just text but video, sensor data, geolocation streams, and structured databases.

In sports, this allows the system to analyze film, detect opponent patterns, and suggest counterplays in real time. In defense, the platform can integrate telemetry from aircraft, radar signatures, and satellite intelligence to support battlefield decisions. In emergency response, the system can correlate emergency call metadata, geolocation signals, and responder availability to optimize response routing—all without needing humans to search through interfaces.

According to TISC co-founder Sabbaghi, the intent is to build software that feels fast, fluent, and native to each user’s domain, rather than behaving like a generic tool that forces users to adapt to it.

How investors are evaluating TISC’s market fit and platform potential in AI decision infrastructure

Investor feedback on TISC has been bullish, particularly on its ability to bridge theoretical advances in machine learning with practical, domain-specific execution. OVO Fund partner Gianfranco Filice stated that he initially considered search a “solved problem” until the TISC team demonstrated how existing AI systems force users to conform to the interface’s logic, leaving intent and nuance underutilized. The gap between a user’s needs and a system’s response is what TISC is aiming to close.

Alex Norman, managing partner at n49p, echoed this sentiment, calling TISC’s vision “category-defining.” He emphasized the startup’s ability to ingest data from any source, understand abstract queries, and deliver precise answers that are usable in the moment—not hours later in a report or dashboard.

For investors focused on AI infrastructure, TISC represents a new kind of enabling layer. Rather than competing with front-end AI tools like OpenAI’s GPT or enterprise dashboards, it complements these systems by offering an embedded retrieval engine that enhances speed, relevance, and domain fidelity. Its real-time focus and API-first design make it attractive for integration across multiple verticals, from sports tech to battlefield software stacks.

What comes next for TISC as it ramps hiring and expands beyond sports use cases

TISC is now scaling its team and partner network to expand into adjacent sectors. The startup plans to onboard more professional and collegiate sports clients while building out use cases in emergency services, public safety, and defense logistics. Its transformer models will continue to be refined to adapt to domain-specific challenges, and new interfaces are in development to support spoken queries, touch-based controls, and API integrations with third-party systems.

According to co-founder Arpan Bhattacharya, the next 12 to 18 months will be focused on doubling down on machine learning research, onboarding strategic pilots, and forming long-term commercial partnerships in critical industries. TISC also intends to demonstrate its technology in new live environments to showcase its agility, accuracy, and ability to handle data streams at scale.

As organizations increasingly look for AI systems that do more than summarize knowledge and instead help guide immediate, high-stakes decisions, TISC is positioning itself as a layer beneath the surface. One that does not replace human intuition but extends it with speed, context, and precision.

What are the key takeaways from The Intelligent Search Company’s emergence from stealth?

• The Intelligent Search Company (TISC) has exited stealth with a US$2.1 million pre-seed round co-led by OVO Fund and n49p, with support from Panache Ventures.

• TISC’s transformer-based platform is designed to enable real-time, high-stakes decision-making by merging retrieval, memory, and reasoning in a single AI system.

• Unlike standard large language model search tools, TISC avoids query lag and irrelevant outputs by predicting intent and retrieving actionable insights across live, multi-modal data streams.

• Early deployments include pilots with a Caribbean national basketball team, NCAA Division 1 games, and the Women’s Premier Basketball Association, which plans to introduce an AI assistant coach powered by TISC in 2026.

• The WPBA is using TISC to centralize scouting, live footage, and performance data to assist coaches and WNBA scouts with real-time evaluations.

• Investors backing TISC believe it addresses a major inefficiency in current AI systems by enabling intuitive search aligned with human intent and domain fluency.

• The platform targets use cases across sports, defense, and emergency response, offering real-time intelligence that augments human decision-makers and AI agents alike.

• TISC’s next phase includes ramping up machine learning hires, expanding pilots, and deepening integrations across adjacent verticals such as public safety and defense operations.