CUSTOMER STORIES
LTM Accelerates AI-Driven Hiring and Transforms Strategic Talent Acquisition with Cortex AI
By unifying its fragmented HR data on Snowflake’s AI Data Cloud, LTM transforms its hiring process to increase onboarding prediction accuracy, boost TCO savings and prevent candidate dropouts.
70%TCO savings achieved from migrating to Snowflake
10xFaster query performance on critical hiring data with Snowpark
Industry
TechnologyLocation
IndiaDriving the future of proactive hiring with AI
Today’s hiring landscape looks nothing like the past. Companies of all sizes and industries have to contend with an increasingly complex global talent market. Hiring today means juggling fierce competition for top tech talent, shifting employee expectations — and the stark reality of high attrition rates. For large enterprises, siloed HR systems, manual processes and delayed insights compound these challenges even further. The costs — and stakes — of reactive decision-making couldn’t be clearer with post-offer dropout rates jumping as high as 30%.
LTM, a global technology consulting and digital solutions company, quickly realized traditional hiring models were no longer enough to stay competitive. To scale hiring strategies and improve candidate retention, LTM recognized that it needed a unified data, analytics and AI platform to bridge the gap between its ambitious business goals and the future of hiring. That’s why LTM turned to Snowflake to consolidate its data cloud platform and help transform hiring with advanced AI and data capabilities.
Story highlights
Improving recruitment predictability and financial efficiency: By achieving 80% accuracy in candidate onboarding predictions, LTM mitigates post-offer dropouts, saving lost productivity and re-recruitment expenses.
Scaling massive operations and performance gains: Thanks to Snowflake’s unified platform, LTM now processes hiring data 10 times faster while reducing the total cost of ownership (TCO) by 70%.
- Accelerating AI/ML innovation and time to market: By leveraging Snowflake Cortex and Snowpark, LTM now develops and deploys AI/ML applications five times faster than before.
Powering cost and time savings with accurate predictions
Prior to Snowflake, LTM’s HR analytics relied on legacy, on-premises systems plagued by data silos, high maintenance costs, poor scalability and slow processing speeds. Candidate predictions were frequently delayed by weeks, inflating recruiting expenses by a magnitude of two to three times during peak hiring seasons.
“The hiring landscape is evolving rapidly, and technology is no longer just an enabler — it is a strategic partner in shaping better candidate experiences and driving measurable outcomes,” says Rajeev Menon, Executive Vice President of Human Resources at LTM. This core insight guided LTM’s decision to migrate to Snowflake’s single, unified platform. Instead of being a cost driver and barrier to scaling, the right data cloud and AI partner could help LTM unlock real-time analytics and elastic scalability, enabling the team to significantly reduce total cost of ownership (TCO) while pursuing ambitious business goals.
LTM implemented its migration to Snowflake in thoughtful phases. First, the team used Snowflake’s Snowconvert tool to minimize downtime while executing a lift-and-shift of critical workloads. Next, LTM modernized its ETL pipelines with zero-ETL connectors to diverse sources such as SAP, SuccessFactors and ATS platforms — maximizing a single source of truth while minimizing cost and complexity. Last but certainly not least, the team deployed ML models via Snowpark to power its end-to-end operations.
LTM’s migration is already paying off. Thanks to its new all-in-one AI and data cloud platform, LTM now uses its AI to analyze historical data from successful and unsuccessful joiners. Thanks to unified data and the power of AI, LTM predicts the probability of candidate onboarding with an 80% accuracy level — 25 to 30 days ahead of the candidate’s start date. Early identification of potential dropouts and proactive strategies to enhance retention are a game changer for hiring teams — and the business’ bottom line.
But LTM’s migration wins go beyond the accuracy of candidate onboarding predictions. Its thoughtful and targeted migration has already enabled 10 times faster query performance, 70% savings in total cost of ownership, and the ability to seamlessly scale performance with hiring volumes.
“By leveraging Snowflake as our data platform, we’ve achieved seamless alignment between technology and business, enabling advanced models like joiner prediction with exceptional accuracy. This milestone is a testament to how innovation and technology converge to redefine and transform recruitment.”
Rajeev Menon
Executive Vice President, Human Resources, LTM
Building a future-ready talent pipeline — and data foundation
To scale its AI-driven hiring, LTM knew it needed to move away from "monolithic" code. Inflexible scripts that choked its legacy systems during high-volume processing were simply not enterprise-ready.
To achieve this, the team leveraged Snowflake’s modular architecture. First, LTM logically separated data stages and utilized virtual warehouses for automatic parallel processing across nodes. This critical shift allows LTM to distribute complex workloads. In real-world terms, the team can now extract vital features from candidate profiles and aggregate dropout histories with unprecedented speed — a gamechanger for making accurate hiring decisions and allocating precious company resources.
Modular data modeling and Snowpark also prevent system bottlenecks, while Dynamic Tables and automated data pipelines streamline the transformation of raw data into consumption-ready datasets. Precision governance, including mandatory object and query tagging, provides full visibility into costs and data lineage while ensuring LTM complies with global regulations like GDPR and India’s DPDP Act. This traceability also helps LTM prove model fairness, secure sensitive candidate data and scale AI safely across more than 80,000 users.
Accelerating AI and hiring innovation with Snowflake Cortex AI
LTM has also supercharged its transformation of talent functions. Thanks to Cortex AI, LTM can now access industry-leading large language models (LLMs) like Snowflake Arctic, Llama 3 and Mistral via a fully managed, serverless environment. By eliminating the need for complex GPU management and model hosting with Cortex AI and integrating Cortex through Canvas.ai, LTM can build and deploy AI applications five times faster than traditional methods. By democratizing insights and intelligence, the team can talk in natural language with employee data and automate high-impact tasks like resume screening and personalized retention insights.
A cornerstone of LTM’s AI innovation journey is its joining probability predictor, an AI/ML model that’s transformed LTM’s hiring strategy from reactive to proactive. By analyzing 12 predictive features from historical data, the model segments candidates into Red, Amber and Green (RAG) categories based on onboarding probability. This empowers LTM’s hiring teams with the critical lead time of 25 to 30 days to implement targeted, timely and data-backed interventions like personalized offers or culture-fit sessions, directly preventing post-offer dropouts that typically cost the company one to three times the candidate’s salary.
By increasing training iterations and expanding historical data, LTM successfully boosted model accuracy to over 80%. Automated tasks now run LTM’s models daily, operationalizing and integrating insights into Power BI dashboards to deliver real-time visibility across the company. Over 700 users across Talent Acquisition, Business, and Operations now leverage this Snowflake-backed solution to drive measurable hiring outcomes.
“Thanks to Snowflake, we can now customize candidate engagement strategies, prevent last-minute disruptions — and most importantly — avoid customer disappointments,” says Menon. With an enterprise-ready data and AI foundation, LTM’s teams are ready for anything the future of hiring holds in store.
