SS. Start a conversation

Principal Product Manager · Oracle Communications

Turning complex operations into governable AI products.

I combine first-hand NOC experience, telecom depth and enterprise product leadership to move network operations from fragmented investigation toward evidence-grounded, human-governed action.

Explore product stories View résumé
17years across telecom, OSS/BSS and SaaS
50+global CSP and enterprise customers
4major product releases owned
0→1network performance product built

Selected product stories

Decisions, evidence and outcomes—not a feature catalogue.

Each story separates what has shipped, what is being validated and what remains part of the product direction.

01
MVP under internal validation

Agentic Operations

From alarm investigation to evidence-grounded, governed action

Defined the product vision, experience and phased autonomy model for an agentic operations environment that brings signals, topology, service impact and operational tools into one accountable workflow.

Agentic AIAIOpsHuman oversightEvaluation
Role
Product vision through MVP
Users
NOC engineers and operations leaders
Evidence
Engineering-built MVP
Current state
Structured testing with internal teams
02
0-to-1 product delivered

Network Performance Management

Closing a strategic capability gap in Unified Assurance

Built a new performance-management module from concept into the portfolio, translating CSP requirements into a reusable product capability and supporting customer pursuits.

0-to-1Network assuranceEnterprise SaaS
Scope
Strategy, discovery and roadmap
Commercial context
New-logo and expansion pursuits
Product context
Unified Assurance
Domain
Fault, performance and topology
03
Product lifecycle leadership

Unified Assurance

Evolving an enterprise assurance platform across customers and releases

Own inbound strategy and lifecycle management across discovery, prioritisation, engineering execution, release readiness, GTM enablement, adoption and post-release feedback.

Platform strategyRoadmapGTMCustomer discovery
Customers
50+ global CSPs and enterprises
Organisation
50+ cross-functional contributors
Releases
4 major releases owned
Growth supported
~US$3M to ~US$8M
04
Validation system designed

AI Evaluation & Governance

Defining what must be true before assistance becomes action

Designed the evaluation dimensions and control boundaries for agentic operations—covering evidence quality, human oversight, tool execution, recovery verification and safe failure.

AI evaluationGuardrailsHuman-in-the-loopEnterprise controls
Scope
Product evaluation framework
Measures
Quality, safety and effort
Controls
RBAC, approval and rollback
Status
Applied to MVP validation

Beyond AIOps

The domain is telecom. The product methods travel.

My current proving ground is network operations, but the underlying product system applies wherever AI must reason across fragmented enterprise context and act through governed tools.

01

Workflow decomposition

Turn expert work into observable decisions, tools, handoffs, failure paths and measurable completion criteria.

Applicable to IT operations, support, care and service delivery
02

Context & orchestration

Bring signals from multiple systems into a traceable situation model before asking AI to recommend or act.

Applicable to cross-system enterprise workflows
03

Evaluation as product design

Define quality, safety, effort and recovery measures around the task—not only around model output.

Applicable to copilots and agentic products
04

Governed execution

Design permissions, approval, rollback and verification into the experience from the beginning.

Applicable to every high-consequence AI workflow
Evidence boundary

The Agentic Operations experience is an engineering-built MVP under structured internal validation. This portfolio shows the product decisions and evaluation system that exist today; production adoption and outcome evidence will be added only when externally supportable.

How I build

Autonomy is a product progression, not a launch switch.

The operating model moves from trustworthy understanding to progressively governed execution.

  1. 01Observe

    Bring events, metrics, logs and topology into situation context.

  2. 02Understand

    Establish service impact, evidence and competing hypotheses.

  3. 03Decide

    Rank next-best actions against risk, permissions and confidence.

  4. 04Act

    Execute through approved tools with oversight and rollback.

  5. 05Verify

    Confirm recovery, capture outcomes and improve the system.

01

Evidence before confidence

Every important conclusion should expose its sources, freshness and limits.

02

Control before autonomy

Permissions, approval and rollback are part of the experience—not backend details.

03

Outcomes before theatre

Measure task completion, operator effort, escalation, recovery and safe failure.

Career journey

Domain depth built from the operations floor upward.

Tata Communications

Enterprise networks, service feasibility, activation and automation.

Vodafone India

Enterprise solutions across connectivity, mobility, cloud and IoT.

Reliance Jio

Enterprise portal UX and orchestration across commercial and operational journeys.

Tata Communications · Tech Mahindra

Business analysis leadership and end-to-end telecom solution design.

Oracle Communications

Unified Assurance strategy, product lifecycle and agentic operations.

About

I have lived the workflow I am now trying to reinvent.

I started close to the network—as an engineer working through service activation and operational systems. That experience still shapes how I build products: begin with the operator’s reality, understand the system around the task, and make progress measurable.

Today I lead product strategy for Oracle Unified Assurance and am focused on how agentic AI can reduce operational friction without giving up evidence, accountability or human control.