Completed an Executive MBA alongside a full-time role, concentrating in Finance and Entrepreneurship & Innovation to complement a deep technical foundation in AI.

NSIT (now Netaji Subhas University of Technology, NSUT) is among India's most reputed institutions for technical education, nationally and internationally recognized for excellence in engineering education and research.

A five-month advanced analytics program (built through a ZS–INSOFE collaboration) covering foundational and advanced statistics, probability, machine learning, and applied data science techniques in R. INSOFE is recognized among the top analytics training institutes in India and is CMU-LTI certified.

Managing cross-functional teams to design, develop, and deploy multi-agentic systems that automate processes and decision-making.

• Leading the automation of a payroll system for a top-5 technology client using AgentOS, PwC's proprietary multi-agentic orchestration engine.

Managed a 12+ member AI team (onshore and offshore) leading AI / Gen AI project delivery, product development, sales, and commercialization within the Health Plan and Provider (HPP) practice.

• Defined enterprise-level AI and Gen AI strategy and transformation roadmaps for business and IT teams across pharma, health plans, and life sciences.
• Led the end-to-end lifecycle and commercialization of 10+ AI / Gen AI powered products delivered across 7+ clients.
• Care & Utilization Management suite (6-member team), where I built and scaled Gen AI products:
  • Post-Call Summarization: generates case-note summaries from case-manager call recordings/transcripts, saving 60–70% of manual effort; deployed for 2 clients.
  • Pre-Call Preparation: distils key insights from disparate sources to prepare care managers for a patient visit/call.
  • Prior Authorization Automation: automates clinical and non-clinical review of prior-authorization requests for utilization-management teams.
• Medicare Advantage Center of Excellence suite (12–15 member team), which I conceptualized and developed:
  • Next Best County: recommends optimal counties for expanding MA plan coverage.
  • Next Best Plan: optimizes MA plan design for county characteristics, SDOH, competitive dynamics, and current design.
  • Plan Document Extractor: automates download and extraction of MA plan details from CMS PBP files, EOCs, etc. into structured data.
  • MA Chat: Agentic-RAG chat over a corpus of MA plan documents; a research tool for MA product designers, achieving up to 80% reduction in manual effort.
  • MA Sales Agent Assistant: helps marketing/sales teams craft optimal, comprehensive responses to inbound prospects.
  • MA Voice of Customer: extracts and curates insights from web/social data, interviews, and call recordings.
  • Broker Engagement Engine: optimizes engagement with external broker teams to lift MA plan sales.
• Implemented VERSO, an omni-channel orchestration product for the specialty pharmacy group of a US health insurer, delivering 250%+ ROI through synergistic orchestration of multi-channel touchpoints for target HCPs.
• Led a 4-member team optimizing relationship management for 80k+ brokers in a US health insurer's Retail (Medicare) division, driving a ~30% increase in MA plan sales and $65MM+ in potential future revenue; the proof of concept won an AI-first strategy & transformation program with the division.
• Led a 6-member team executing breadth, depth, and attrition-driver analyses for key HIV drugs for a British pharma company.
• Led a 5-member team designing a Risk Stratification offering under Population Health Management.

Represented and led the 35+ member Artificial Intelligence (AI) Track in the New Delhi office, driving projects and product development across Health Plan & Provider (HPP) and Non-Healthcare (NHC) Analytics ventures.

• Built and scaled high-performance AI teams: the New Delhi AI Track to 35+ members and the NHC Analytics venture to ~15 members.
• Led two teams (7 AI members total) designing omni-channel commercialization models for HCPs across two pharma companies, spanning patient-event prediction, influence mapping, HCP segmentation, channel/topic affinity, and patient-pathway analyses, deployed across six drugs/indications. Won ZS Project Champion (Client Impact, Innovation of the Year) in 2021.
• Led an 8-member team building the patented AI-Guided Selling product for B2B inside-sales teams, a cloud-agnostic, hybrid-SaaS suite of 10+ predictive and prescriptive models (churn/leakage, lifetime value, upsell/cross-sell/net-new-sell, purchase value, relationship management, conversion optimization, and CHI forecasting) built on an AutoFeaturization + AutoML pipeline with an explanation-model framework for low-inertia adoption.
• Contributed as an individual contributor within the Artificial Intelligence Center of Excellence (AI CoE), focused on computer vision in medical imaging and AI-driven product discovery.
• Developed in-house AI/ML system DevOps capability to automate ML lifecycle management and deployment.

Led a 12+ member team building a portfolio of ML/AI products for a Big-5 technology company, and a 6-member team creating AI products in the Real-World Evidence (RWE) & Clinical Trial Optimization (CTO) domain under R&D Excellence.

• Led a 10+ member team enhancing a multi-model ecosystem for a top-5 technology company, scaling identification from ~7M prospects across 100+ countries to prioritization of ~700K leads across sales channels, and evolving it from a 2-model to an 8-model ecosystem across 7+ release cycles.
  • Delivered ~$70M in aggregate annual revenue growth (~$50M via a purchase-value model, ~$13M via a lead-prioritization model, and ~$8M by integrating four new data sources); achieved AUROC of 84% on a new partner-channel propensity model. Won ZS Project Champion (Innovation of the Year) in 2019.
• Developed propensity models advancing a client's GTM strategy across 18K+ accounts, achieving AUROC of 78% in predicting propensity to buy.
• Developed an algorithm predicting a specific medical diagnosis within six months from patient history and event data, achieving AUPRC of 66% via a 1-D CNN-RNN hybrid to identify 25,000+ naïve patients/year and expose a $150MM/year opportunity from highly imbalanced EHR data (~2% positive class).
• Designed an AI-enabled automated schema-matching process converting raw clinical-trial datasets of disparate schemas to a standard format, achieving 87%+ accuracy at ~2 mins/dataset (an 11-point improvement over the client's in-house team), driving $8MM in direct cost savings and attracting $800K of new project work. Techniques: shingle-phrase mapping, semantic lookup dictionary, sentence2vec, WordNet, SVM, XGBoost, neural networks. Won ZS Project Champion (Innovation of the Year) in 2018; showcased at SCDM 2018 (world's top clinical data management conference).

Among the first cohort (~15 in the US, ~5 in India) laterally promoted to the Advanced Data Science Track through a company-wide shortlisting process.

• Completed the 5-month Advanced Analytics Learning Program (ZS–INSOFE collaboration).
• Authored 12+ hours of training material on data handling, machine learning, and visualization in R, and mentored ZS New Delhi employees and trainers.
• Ran 5+ advanced analytics sessions and competitions for 250+ employees as a key member of AlgoRhythms, an internal analytics-upskilling initiative.
• Developed and facilitated case studies and hackathons for 700+ participants across premier institutes with the data science recruitment team.
• Built a discrepancy-detection system on CDISC SDTM data (R-Shiny dashboard automating ingestion, integration, anomaly detection, and report synthesis) achieving 10X+ reduction in CDR-team review effort, with an active-learning loop from reviewer feedback; showcased at SCOPE Summit 2018.
• Built a search-and-optimization process to rebalance inclusion/exclusion criteria for clinical-trial enrollment (PoC).
• Created a replicable, customizable disease-KPI dashboard template for a Swiss pharma company (R, Shiny, htmlwidgets, HTML/CSS/JS; Impala, SQL).

Worked on projects across the managed-care practice area in the US healthcare market.

• Created a payer-contracting tool built on patient-level forecast modeling that broke new ground in forecasting patient-level dynamics in the specialty therapeutic market, earning appreciation from ZS principals and the client.
• Developed portfolio-rebating constructs helping a UK pharma company improve share and access in the diabetes market while preserving prices of its latest drugs.
• Formulated a market-access strategy to optimize a US pharma company's therapeutic-drug performance.

Consumer Packaged Goods & Quantitative Research (2013 – 2014); Technology, Media & Telecommunication (2012 – 2013). Projects spanned Telecom, FMCG, Health, Tourism, Lifestyle, and Pharmaceuticals.

• Led a 3-member team and supervised a 26-member vendor team for a brand-tracking study for an African nation's tourism ministry, managing questionnaire fielding, data quality, and output; placed on-site in a West African country to oversee interviewer training.
• Assessed the impact of a global TV ad campaign, building an estimation model to measure a concept's effectiveness in drawing leisure visitors, benchmarked against the existing concept via quantitative survey.
• Streamlined tracking of email promotional activity across 30+ distribution channels for a US lifestyle company, automating the analysis and reducing reporting latency.
• First Runner-Up, Training Case Study Competition: Evolution of Smart Grid Technologies in Developing Countries.
• Tools: MS Excel, PowerPoint, R, SPSS, MS Access, MarketSight, SQL.

• Conceptualized and led a 15+ member team to organize InnoCon, the first NSIT Research Convention, during Innovision'11, NSIT's inter-collegiate technical festival. Executed a two-day paper-presentation and Industry–Faculty–Student interaction attended by 200+ students; engaged a leading telecom company's VP and eminent research professors, drawing critical appreciation.

• Organized an IEEE Delhi Section-level Industry–Academia "Seminar on Telecommunication" (70+ attendees) and initiated the practice of video-archiving IEEE NSIT events.

• Popularized the platform on campus through workshops, poster campaigns, and event-awareness drives.

• Inducted and mentored 50+ students into IEEE NSIT (highest by any individual).
• Co-organized a 5-day Embedded Systems Workshop (75+ attendees) and an industrial trip to NDPL (30+ attendees).
• Handled final editing and publication of IEEE NSIT's in-house publications YUGMA and MAGNOLIA.

• Studied GSM networks, AT commands, and the worldwide collection, exchange, and dissemination of meteorological data.
• Worked on TRANSMET / Automatic Message Switching System (AMSS), the central node of IMD's national and international communications.

Mentors: Dr. N. K. Pangasa (Scientist E, Telecom); Mr. Sourav Adhikary (Project Coordinator).

• Solar Updraft Tower: designed and assembled a prototype and analyzed its performance and commercial application; iterative experimentation yielded an 80% improvement in performance.
• Dye-Sensitized Solar Cells (Grätzel cell): fabricated cells and studied advantages versus conventional cells, increasing generation by 200% (0.1 → 0.3 V/cm²).
• Designed and built a working Foot-Step Electric Converter transforming footfall impacts into electrical energy.

• Predicting pulmonary embolism (PE) and related conditions from 3D DICOM imagery and embedded metadata, working with a rare positive class (<2%) across 1.5 TB+ of medical images from ~8,000 patients (Radiological Society of North America dataset).

• Designed and simulated a fully customizable, high-data-rate duplex indoor OWC system in MATLAB/Simulink, using MIMO and Wavelength Division Multiplexing (WDM) to boost throughput, demonstrating the viability of LEDs for simultaneous lighting and communication.

• An 8085-microprocessor-based system (PCB designed and built in-house, programmed in Assembly) that adapts traffic-light timing to real-time conditions, a low-cost, economically deployable alternative to image-processing solutions.

• Investigated orbits of the Sun–Earth–Moon system as a circular-restricted three-body problem (CRTBP), simulating trajectories in MATLAB via nonlinear second-order differential equations.

• Assembled a 3.3 V/5 V dual-voltage, 1 A power supply (0.5 A foldback limiting) from a 220 V, 50 Hz source using an LM723 regulator, with current and foldback limiting to protect the circuit.

• Implemented Conway's Game of Life on a Nokia 3310 LCD interfaced with an ATmega8 microcontroller.

• A series of independent simulations, including: a PWM-driven DC motor model (Simulink); a triangle-wave generator (SimElectronics); solid-shape generation via random numbers; Brownian motion of N particles; a Hamming-code model; Lambertian LED emission in indoor environments; and introductory image-processing (picture merging).