DPDP Act readiness: a practical checklist for IT and engineering teams
What India's Digital Personal Data Protection Act and the DPDP Rules 2025 mean for systems, logs, consent and breach response, with a phased checklist.
Data engineering and AI that people trust enough to act on
We design data platforms, build analytics your teams actually use, and put machine learning and generative AI into production with the controls needed to keep them there.
What you can expect
- A prioritized use case backlog with value estimates
- Pipelines with tests and alerts, not just scripts
- Models with monitoring and a retraining plan
- Clear ownership for every critical dataset
Overview
Modern data platforms, analytics, machine learning and generative AI, with the governance to keep them trustworthy.
A model is only as useful as the data under it and the process around it. We start by scoring use cases on value and feasibility, then build the pipelines, quality checks and ownership that let results hold up in front of a finance review or an auditor.
For generative AI, that means retrieval grounded in approved sources, evaluation sets written with your experts, guardrails on inputs and outputs, and cost monitoring from the first week.
Our team works from Pune. We can work onsite in and around the city, and remotely for organizations across India and abroad.
Capabilities
What our data and AI team does
- 01
Data platform engineering
Lakehouse and warehouse platforms built for scale, cost and control.
- Microsoft Fabric, Databricks, Snowflake, BigQuery and Redshift
- Batch and streaming pipelines
- Migration from SQL Server, Oracle, Teradata and Netezza
- Data quality checks and observability
- 02
Analytics and BI
Metrics that mean the same thing in every meeting.
- KPI definitions and semantic models
- Power BI, Tableau and Looker dashboards
- Self-service analytics enablement
- Report rationalization
- 03
Machine learning and MLOps
Models that are trained, deployed and monitored like any other software.
- Forecasting, anomaly detection, churn and recommendations
- Feature stores and model registries
- Automated training, deployment and monitoring
- Explainability and drift tracking
- 04
Generative AI and agents
Assistants and workflows grounded in your own knowledge.
- Use case discovery and business case
- Retrieval-augmented generation on your documents
- Copilots for service, sales, HR and IT
- Evaluation, guardrails and cost controls
- 05
AI data services
Human-in-the-loop work that makes training data usable.
- Image, video, text and speech annotation
- Data collection and curation
- Human review of model output
- Multilingual datasets, including Indian languages
- 06
Data governance and privacy
Ownership, lineage and access rules that hold up to scrutiny.
- Catalog, lineage and data ownership
- Access policies and masking
- Data handling aligned with the DPDP Act and GDPR
- Retention and deletion rules
Approach
How the work runs, step by step
Each step ends with something you can review, so decisions are made on evidence rather than status updates.
Step 01
Assess
Data estate inventory, quality review and use case scoring.
You get Prioritized backlog
Step 02
Prove
A time-boxed proof of value on real data.
You get Go or no-go evidence
Step 03
Build
Platform, pipelines and models delivered with CI/CD.
You get Tested pipelines
Step 04
Adopt
Training, documentation and change support for users.
You get Adoption plan
Step 05
Operate
Monitoring, cost optimization and model retraining.
You get Runbook and dashboards
Platforms and tools
Technology we work with
Names are trademarks of their respective owners, used only to describe technologies our teams use. No partnership or endorsement is implied. See the main platforms we work with.
- Microsoft Fabric
- Azure Synapse Analytics
- Databricks
- Snowflake
- Google BigQuery
- Amazon Redshift
- dbt
- Apache Kafka
- Apache Airflow
- Power BI
- Tableau
- Azure OpenAI
- Amazon Bedrock
- Google Vertex AI
- MLflow
Industries
Where this work matters most
- Retail & E-commerceCommerce platforms, personalization, store systems and workforce management.
- Banking, Financial Services & InsuranceDigital journeys, core integration, risk analytics and security for BFSI.
- Healthcare & Life SciencesPatient and clinician apps, interoperability, clinical data and secure cloud.
- Manufacturing & IndustrialConnected plants, ERP and MES integration, supply chain and quality analytics.
- Telecom & MediaNetwork operations, OSS and BSS integration, content platforms and analytics.
- Travel, Transport & HospitalityBooking journeys, guest experience, pricing analytics and service operations.
Questions
Data & AI: frequently asked questions
Where should we start with generative AI?
With two or three use cases that have a clear owner, accessible data and a measurable outcome. We run a short discovery, score each use case on value and feasibility, and build a proof of value before any wider rollout.
Can you migrate our data warehouse to the cloud?
Yes. We inventory the estate, score each object for complexity, design the target architecture and migrate in waves, reconciling data at every step.
How do you keep AI outputs reliable?
Through evaluation sets built with your subject experts, retrieval grounded in approved sources, guardrails on inputs and outputs, human review where the stakes are high, and monitoring after launch.
Will our data leave India?
Only if your architecture and contracts allow it. We can design for Indian cloud regions and document any cross-border flows for DPDP Act and GDPR purposes.
Often combined with
Related services
- 03Cloud ServicesCloud migration, modernization, DevOps and FinOps across AWS, Microsoft Azure and Google Cloud.
- 01Enterprise ApplicationsImplement, integrate and support ERP, CRM and HCM platforms, including Oracle, SAP, Salesforce, Workday, UKG and Microsoft Dynamics 365.
- 05CybersecuritySecurity testing, zero trust, identity, detection and response, and compliance programs you can measure.
Further reading
Generative AI use cases that survive the pilot
Why many generative AI pilots stall, and the selection criteria, evaluation and guardrails that help the good ones reach production.
Talk to our data and AI team.
Tell us about your systems, timelines and constraints. We will come back with questions, options and a suggested first step.