Multimodal Retrieval and RAG
Researching tri-modal retrieval where text, image, and table evidence can be retrieved, aligned, evaluated, and used in downstream reasoning.
Doctoral research at SIT-NVIDIA Joint AI CenterAI research, data products, and leadership
SK Ahammad Fahad
AI researcher and data science leader building multimodal retrieval systems, RAG architectures, and production data products.
Researcher at the SIT-NVIDIA Joint AI Center, Doctor of Engineering candidate, and former data science leader for national-scale retail intelligence systems.
Professional profile
Based in SingaporeAI research + data science leadership
ResearchSIT-NVIDIA Joint AI Center
SystemsPricing and forecasting
EvidencePublications and credentials
Leadership7+ yrs
Data science teams and systems
Publications11
217+ citations
Credentials156
Verified learning portfolio
ResearchSNAIC
SIT-NVIDIA Joint AI Center
Start here
Choose the evidence path that fits your reason for visiting.
The full site remains available, but the first layer now guides collaborators, hiring leaders, technical reviewers, and credential verifiers to the right proof.
Current focus
A profile built around research, systems, and measurable decision support.
Production Data Science Systems
Built pricing, demand, profitability, inventory, and market intelligence systems for national-scale retail and e-commerce operations.
Dynamic pricing and market intelligence projectsResearch-Backed Leadership
Combines research discipline, data product thinking, and team-building experience to move from problem framing to measurable systems.
7+ years leadership, 11 publicationsSelected case studies
Applied work where research discipline meets operational ownership.
A smaller set of stronger examples replaces a wall of equally weighted cards.
Cross-Modal Information Retrieval for Healthcare (Tri-Modal)
Built a tri-modal information retrieval system for healthcare data (image ↔ text ↔ table).
- Role
- Problem framing, system design, cross-modal alignment strategy, evaluation/benchmark planning.
- Impact
- Working research-grade system and evaluation battery enabling measurable iteration on retrieval quality.
Dynamic Pricing Decision Engine (Goal-Driven Pricing)
Implemented goal-aware pricing logic (acquisition vs margin vs revenue) powered by market signals.
- Role
- Strategy rules, objective modes, guardrails, and integration with business workflows.
- Impact
- Faster market response and better control over acquisition pushes while protecting margin.
Market Price Intelligence Ecosystem (Multi-Source Collection)
Built a daily market price intelligence pipeline from digital, traditional, and vendor sources.
- Role
- End-to-end acquisition design, operational process, data quality and freshness standards.
- Impact
- Reliable market pricing signals powering faster and more confident pricing decisions.
Research proof
Selected publications
Peer-reviewed work and surveys that connect AI methods, NLP, business intelligence, and multimodal systems.
Advancements and applications of multimodal large language models: Integration, challenges, and future directions
2025 · AI-Driven: Social Media Analytics and Cybersecurity
Inflectional Review of Deep Learning on Natural Language Processing
2018 · IEEE International Conference on Smart Computing and Electronic Enterprise
Business intelligence through big data analytics, data mining and machine learning
2019 · Data Management, Analytics and Innovation (ICDMAI)
Learning proof
Credential atlas
156 certifications become stronger when shown as a structured skill map rather than a long first-screen archive.
Machine Learning and AIDeep Learning and Neural NetworksData Science and StatisticsBusiness IntelligenceCloud and Data EngineeringResearch Ethics
156Total credentials
93.6%Verification coverage
Research rigor, business ownership, and continuous learning in one profile.
Explore the research program, delivery portfolio, publications, and credential evidence behind an AI and data science profile built for both scholarship and applied systems.