Nov 24, 2025
AI ServeSmart Digest
Insights at the intersection of AI and enterprise strategy, helping leaders turn innovation into impact.
Welcome to the AI ServeSmart Digest, designed for leaders who are shaping the future with AI. Each month, we bring you sharp insights and real-world stories on how applied AI is solving today’s toughest business challenges, creating measurable impact, and opening new growth opportunities. Think of it as your executive lens on what’s next in enterprise AI.
AICS embodies the very essence of Fractal: the fusion of human ingenuity and AI to create lasting impact. Over the past year, the AICS team has demonstrated how agentic AI can unlock exponential value by transforming decision-making across the enterprise. As we enter the next chapter, our focus is on scaling these innovations responsibly and empowering every team to harness AI as a trusted partner for sustainable growth.
Pranay Agarwal
Chief Executive Officer, Fractal
INSIGHT
Transforming telecom operations with Agentic AI
Telecom organizations today operate in a landscape marked by immense data volumes, complex network infrastructures, and constantly evolving customer expectations. Despite significant technology investments, many operators continue to face persistent challenges, including unpredictable network congestion, inefficient resource allocation, and rising maintenance costs.
The Cogentiq Network Intelligence is designed to change that. Built as a multi-agentic generative AI platform, it enables telecom enterprises to orchestrate insights, automate decisions, and optimize outcomes across diverse operational domains.
Why your GenAI success depends on what happens before chunking
Consider this scenario: Your organization has made significant investments in an advanced Retrieval-Augmented Generation (RAG) system. You possess state-of-the-art embedding models, sophisticated retrieval algorithms, and ample GPU resources that satisfy any data scientist. Nevertheless, your Generative AI applications continue to deliver subpar responses, overlook crucial contextual information, and cause user frustration due to irrelevant outputs.
OTHER READS
Agentic diagram parsing: Turning visual logic into enterprise intelligence
Modern enterprises run on visual artifacts, flowcharts, SOP diagrams, customer-journey maps, escalation flows, and process schematics. These diagrams hold critical operational logic yet remain invisible primarily to automation systems.
Seeing the invisible: Observability in the age of agentic LLMs
As language models evolve from passive text generators into dynamic, autonomous agents capable of reasoning and decision-making, a new challenge emerges, understanding how and why these systems behave the way they do.
Toward scalable personalization of language models
Large Language Models (LLMs) have reached enterprise scale, but not enterprise relevance. While models like GPT, LLaMA, and Mistral excel at general reasoning, they fall short when applied to specialized domains or individual user contexts.
Contributors
Ayushi Singh Chhetri
Senior Data Scientist
Vishnu KT
Manager
Karan Samani
Lead Data Scientist
Abhijit Guha
Client Partner
Sumukh Bhalchandra Sule
Data Scientist
Chandramauli Chaudhuri
Client Partner
Soumo Chakraborty
Principal Architecture
Anindya Sengupta
Client Partner
Sujit Shahir
Principal Data Scientist
Prosenjit Banerjee
Principal Data Scientist
Anik Chakraborty
Principal Data Scientist
Tanmay Garg
Lead Data Scientist
Mandar Patil
Lead Data Scientist
Parul Chaudhary
Data Scientist
Triparna Chatterjee
Associate
