Fidelity Investments

AI Leader - Intelligent Automation

Leading Data science & AI strategy for Operations Analytics supporting Fidelity's Workplace Investing and Healthcare business units.

Director -  Machine Learning

Built Nextgen Analytical products & services for Operations Analytics

1. Forecasting at Scale - Built home grown advanced forecasting product powered by AI 

	a. Forecast univariate time series data with 20+ learning algorithms
        b. 10x faster than  traditional  auto forecast   engines 
	c. Forecast multivariate time series data with causal inference of predictors influence 
	d. Time series anomaly/event detection and event annotation 
	e. Scaled the product to forecast all back office work items  of WI & HCG business units
	f. Classical / custom decomposition of time series data 
        g. Time-series insights writer to provide  actionable insights 
        h. Real time forecasting engine that trains/scores & gives the best model in few seconds to minutes

2.Tactical Plan Recommendation engine for Operations (AI to effectively manage the Demand - Supply)
Developed Recommendation engine that optimises (Productivity / Overtime / Utilisation) for operations associates & managers  which would suggest tactical plan for the upcoming week.

3. SCOBOT  ( AI to estimate the cost/hours & resources for technology projects)
Developed an AI based scoping engine that estimates hours/ billing & required role resource pools for the implementation of a tech project.The engine is built using adaptive binning algorithms and discretisation and does not require re-training of model.

4. Architected & led the development of  Multi Year / Long Range Planning Product - an Analytical platform that drives Hiring / Budget decisions) of Workplace solutions & Health care business units of Fidelity.

5. Q-GO (Personalized Content Recommendation Engine) Built Q-Go a voice based search bot to identify the relevant content for phone representatives and push the content real time to representative’s desktop.

Senior Manager -  Data Science

Citi

Citigroup - Global Analytics & Insights

Led multiple projects in the below areas
Marketing Analytics, Customer Segmentation, Customer insights for acquisition, retention, profitability and loyalty (Citi Commercial Bank), Global Technology Resource Strategy, Vendor Management Analytics, Cost optimisation for tech spend of Citi technology BUs.

Analytics Manager

🚀 I'm Yashwant Reddy (Sai), a seasoned AI leader with a proven track record in building high ROI AI solutions. As Director of Machine Learning at Fidelity Investments, I lead the development of cutting-edge AI solutions to enhance customer experiences in call centers. 

💼 With 13+ years of experience in the field, I've honed my technical skills in:

• Artificial Intelligence (AI)
• Machine Learning
• Natural Language Processing (NLP)
• Generative AI
• Large Language Models (LLM)

🔑 Key Leadership Skills:
• Leading top-tier AI teams
• Offering strategic direction 
• Identifying high ROI AI opportunities 

✨ I'm driven by the potential of AI to create a smarter, more efficient world. I believe in continuously learning and adapting, ensuring that the AI solutions we develop are not just advanced but also ethically and socially responsible.

🎓 My academic background includes a Master's degree in Computer Science from the National Institute of Technology Karnataka and an executive education on AI Applications for Growth from Kellogg's Executive Education.

Disclaimer: The views, thoughts, and opinions expressed in my posts on this platform are entirely my own and should not be attributed to, nor do they represent the official stance, strategies, or opinions of my employer or any other organization I am affiliated with. All content is provided for informational purposes only and is not intended as official communication on behalf of my employer.

Yashwant R.

Recently completed Hamel Husain and Shreya Shankar "AI Evals" course on Maven. If you're building AI products and moving beyond proof-of-concepts, this is essential learning.

AI Evals is both "art and science". You need both to ship production systems that deliver consistent value.

Top 3 Learnings from the bootcamp

1. Error Analysis > Generic Benchmarks : Instead of asking "how does this perform overall?" ask "what specific patterns of failure exist?" Systematic error analysis reveals exactly where your system breaks, in RAG retrieval, multi-step reasoning, or domain edge cases. This creates a clear engineering roadmap and eliminates guesswork about where to invest effort.

2. LLM-as-a-Judge Requires Rigor : Using peer LLM to judge outputs isn't enough on its own. The course teaches an iterative process: align with domain experts, build calibration sets, measure reliability, and validate continuously. Continuous calibration is more important than continuous deployment.

3. Know When NOT to Write an Eval : Not everything needs an eval. The key is distinguishing high-leverage evals that unlock rapid iteration from low-ROI vanity metrics. This judgment, understanding trade-offs between eval coverage, velocity, and cost lets you to ship products quickly with “Honest uncertainty”

Essential for anyone responsible for delivering a high-quality AI Product.

What evaluation challenges are you facing? I'd love to compare notes. 
Course link in comments.

Sai Yashwant

A key takeaway from conversations with global AI leaders at Tech Circle 2025: success with agentic AI begins with the purposeful design of human-AI collaboration. When the architecture is intentional, humans and autonomous systems complement each other at their highest leverage.

Thank you Sandhya Ramachandran Arun and the Wipro Intelligence team for organizing Tech Circle 2025.

Grateful to be part of this community as we figure out what responsible enterprise autonomy actually looks like in practice.

#agentiAI #entepriseAI

CTO @ Wipro | Driving Strategic Innovation, Frontier Tech R&D, AI Transformation, Digital Experience, Products and Platforms, Open Innovation Communities

View Sandhya Ramachandran Arun’s profile

AI leaders, researchers, and strategists from global enterprises and research institutes, came together at the Tech Circles at our Kodathi Innovation Lab in Bengaluru.  The theme, "Agentic Enterprises - from Automation to Autonomy", evoked high energy discussions, debates and constructive dialogues. While emphasising the need for pragmatism and the importance of de-risking AI adoption, there was consensus that AI solutions must be both responsible and scalable by design. AI forward enterprises are evolving with both Responsibility and Autonomy edited into their DNA. Most importantly, we concluded that  an AI-First approach to enterprise transformation must be human centric. A big thank you to all our speakers, guests and organizers! #WiproInnovationNetwork #WiproIntelligence 

A big shout out and heart felt thanks to:

The speakers and panelists for sharing your invaluable perspective - Sai Yashwant,Sundar R Nagalingam,Ankit Bose,Teena Bhasin, Soumitra Saha, Karthik Rameshkumar, Shivakumar KR, Ambedkar Dukkipati, Kishore Kumar Ambiti,  Ajay Bhaskar, Brian Rizman and Pushpa Ramachandran

The Organizers -Jessy Alina Konatham, Varun Dube, Padmaja Sreenivas Prachi, Manish Kumar, Hari Shetty, Ranjita Ghosh

Take it away at Sydney, Ketan Panchal, Dr Arash Dadras, Gautam Nadkarni, and Kiran Minnasandram...

taplio