Sep 2023 – Sep 2025 · 2 yrs · On-site

• Deputed as Centre Head at the India–Solomon Islands CEIT, leading digital transformation at SINU in collaboration with the Government of Solomon Islands, under C-DAC (ICD Division), MEA & MeitY, Govt. of India.
• 🧭 Leadership & delivery: Directed CEIT roadmap (platforms, training, stakeholder alignment), vendor coordination, and budgets; ensured high availability across academic cycles for teaching/admin systems.
• 🗂️ Platform modernization: Oversaw university systems—grading/workflows, HRMS, policy & knowledge portals—standardizing data models, access controls (RBAC), and documentation for auditability.
• 🤝 Change management: Ran discovery → BRD/FRD → UAT → release; set up release notes, runbooks, and issue triage so teams could ship safely during teaching periods.
• 🎓 AI enablement (practical & safe): Designed intro AI workshops for faculty/staff (prompt basics, privacy, responsible use) and published simple playbooks for day-to-day tasks (summaries, drafting, form helpers).
• 📚 Policy Q&A pilot (RAG): Built an internal policy/procedure assistant with FastAPI + LangChain + PostgreSQL/pgvector; grounded answers with source citations, added input/output checks and basic rate limits.
• 📈 Observability & cost hygiene: Added structured logging for latency and token usage, lightweight cost dashboards, and safe defaults (timeouts, retries, caching) to keep pilots reliable and predictable.
• 🛡️ Governance: Introduced PII redaction, approval workflows for new data sources, and simple audit trails; drafted a DPIA-style checklist to guide future AI use on campus.
• Selected outcomes:
• 1. Faster access to policies/procedures and fewer repetitive support queries.
• 2. Teams onboarded to safe, governed AI usage with clear playbooks and fallbacks.
• FastAPI • LangChain • PostgreSQL/pgvector • Streamlit • Python • REST • RBAC • Logging/Tracing • Cost/Latency dashboards • Prompt basics & evaluation lite • Privacy/PII • Audit trails • BRD/FRD • UAT • Higher-education IT

Oct 2022 – Sep 2023 · 11 mos · On-site

• 🧪 Multimodal AI program (vision + biosignals): Led end-to-end R&D across computer vision and biosignals (EEG/ECG/GSR), from data collection and labeling to model training, evaluation, and dashboarding.
• 🎧 Data collection pipelines: Set up field and lab capture using Emotiv EEG headsets (multi-channel), ECG/PPG modules, and webcam/video streams; built Python scripts for synchronized acquisition, metadata, and consent logs.
• 🧹 Pre-processing & feature pipelines: Artifact removal (band-pass, notch), ICA/ASR for EEG, peak detection for ECG; epoching & normalization; generated STFT spectrograms/scalograms and classic features (bandpower, HRV).
• 🧠 EEG/ECG modeling (CNN/CNN-LSTM): Trained 1D-CNN and CNN-LSTM models on EEG/ECG windows; experimented with shallow ConvNet-style architectures and attention pooling on spectrograms for state/condition classification.
• 👁️ Computer vision models: Built ResNet/MobileNet baselines in PyTorch for engagement/activity cues; used OpenCV + MediaPipe for face/pose landmarks; tried YOLO for event detection where appropriate.
• ⚙️ Training ops & experimentation: PyTorch Lightning, TensorBoard/Weights & Biases for experiments, Optuna for HPO; stratified cross-validation (by subject/session) and subject-held-out tests for generalization.
• 🖥️ Dashboards & serving: Built Streamlit/Dash dashboards for live monitoring and review; exported models; packaged FastAPI inference services with basic auth, logging, and audit trails.
• 🧪 Validation discipline: Created golden test sets, confusion matrices, ROC/AUC/F1 tracking; checked session/device drift and reproducibility (seed control, deterministic dataloaders).
• 👩‍🏫 Training & enablement: Ran hands-on sessions for analysts and partners on pipeline usage, model interpretation, and safe/ethical application of AI in education/health contexts.
• PyTorch • PyTorch Lightning • TensorFlow/Keras • scikit-learn • Optuna • MLflow/W&B • OpenCV • MediaPipe • MNE-Python • neurokit2 • biosppy

Oct 2019 – Oct 2022 · 3 yrs · On-site

• Deputed as Master Trainer cum Center Head at the India-Niue Centre for Excellence in IT, Alofi, Niue driving digital transformation at Niue in collaboration with the Government of Niue Island, under C-DAC's ICD Division, managed by MEA and MeitY, Government of India.
• 🛠️ Service delivery & infrastructure: Led ICT operations for training labs and administrative systems; sustained 98%+ uptime via preventive maintenance, structured ticketing, and change control.
• 🗄️ Digital transformation of records: Digitized 1,500+ employee files (service history, training, appraisals) with governed fields and access controls; reduced retrieval times by ~70% and improved PSC workflows.
• 📊 Decision dashboards: Designed Power BI dashboards (participation, completion, department engagement) for the Public Service Commission—improving reporting efficiency by ~40% and enabling timely, data-driven decisions.
• 🎓 Capacity building: Delivered 200+ hands-on ICT programs to 400+ civil servants (Microsoft 365, data analysis, IT fundamentals, cybersecurity). Authored reusable guides and checklists to standardize practice.
• 🧭 Process & governance: Produced monthly/quarterly/annual KPI reports on usage, training impact, and reliability; established access controls, backup routines, and documentation aligned to government expectations.
• 🤖 Early AI enablement (intro): Ran intro sessions on practical AI/automation (prompt basics, privacy, responsible use) and set up simple drafting helpers (FAQ/email templates) to reduce repetitive work—no sensitive data, clear guardrails.
• Stack: 🐍 Python (scripts) • 🌐 REST basics • 🐘 PostgreSQL • 📈 Power BI • 🖥️ Windows/Linux admin • 🎫 Ticketing/Helpdesk • 🔁 ETL • 🔐 RBAC/Access Control • 💾 Backups/DR • 📌 KPI Reporting • 🧩 Intro AI training • 🛡️ Privacy & safe use • 🏫 Education/Public Sector

Aug 2019 – Oct 2019 · 2 mos · On-site

May 2023 – May 2025 · 2 yrs · India