AI Application Support

Promaynov Advisory Services Pvt. Ltd; Bengaluru; On-site; INR 1,200,000 - 1,800,000; Full time

Job summary

Promaynov Advisory Services Pvt. Ltd. seeks an experienced L2 Support Engineer to provide advanced troubleshooting for AI-based applications. You will triage, diagnose, and resolve incidents escalated from L1, distinguishing functional queries from technical issues, and elevating to L3 with complete context.

Focus on building knowledge management, runbooks, and observability improvements to support rapid scale from 20 to 100+ apps. Strong collaboration with L1/L3 and product teams is essential.

Qualifications

• 4+ years of production support/operational experience in cloud-native environments.
• Experience with AI-enabled applications and integration patterns.
• Strong triage, problem-solving and diagnostic skills.
• Excellent documentation and communication abilities.

Responsibilities

• Pick up escalations from L1 Service Desk and own resolutions or escalation.
• Validate impact, urgency, priority; ensure correct categorization and assignment.
• Perform structured triage to classify issues as technical incidents or other.
• Apply runbooks and document actions in ServiceNow.

Skills

AI platforms

Cloud fundamentals

Troubleshooting

Documentation

Tools

ServiceNow

Azure Monitor

OpenAI

Job description

Experience : 4-6 years

Role purpose : Provide advanced Level 2 (L2) support for AI-based applications by performing deep triage, troubleshooting, and resolution of incidents and service requests escalated from L1 Service Desk. Distinguish functional queries from technical incidents, restore service quickly, and elevate to L3 Product Support (Advisory UK service lines) with complete diagnostic context. Strengthen operational readiness for rapid scale (20 apps to 100+ apps) through knowledge management, documentation, and observability/reliability improvements.

Key responsibilities (what you will do)

• Pick up escalations from L1 Service Desk and take ownership through resolution or appropriate escalation.
• Validate and confirm impact, urgency, and priority, ensuring correct categorization and assignment.
• Perform structured triage to classify issues as:
• Technical incidents (errors, outages, latency, integration failures, platform degradation).
• Apply runbooks/diagnostic playbooks to resolve issues within L2 scope and document all actions in ServiceNow.

2) Technical Troubleshooting Across AI + Cloud Stack

• AI platforms: OpenAI / Azure AI patterns such as authentication errors, throttling (rate limits), token/context issues, deployment/model availability, safety/content filtering impacts, latency/timeouts.
• Application dependencies: integrations, data sources, retrieval pipelines (if applicable), middleware, and downstream service availability.

3) High-Quality Escalation to L3 (Advisory UK Product Support)

When L2 resolution is not possible, upscale to L3 with a complete evidence package to minimize back-and-forth:

• reproduction steps and expected vs actual behaviour
• logs/metrics/traces (e.g., App Insights/Azure Monitor where applicable)
• recent changes/releases, interim mitigations applied, and L2 hypothesis
• sanitized payload samples (compliant with data handling rules)

4) Knowledge Management & Documentation (Deflection + Speed-to-Resolve)

• Create and maintain Support Manuals, Runbooks, SOPs, and Troubleshooting Guides for each supported AI application.
• Write and curate Knowledge Articles in ServiceNow to enable L1 deflection and consistent resolution.
• Maintain “Known Errors / Known Issues” records and update them based on recurring patterns and fixes.

5) Observability, Reliability & Operational Readiness

• Identify monitoring gaps and recommend improvements (alerts/dashboards/logging standards) to reduce MTTR.
• Support go-live readiness for new AI apps by ensuring:
• support model and escalation paths are defined
• runbooks and known dependencies are documented
• logging/telemetry is sufficient for L2 diagnosis
• common failure modes and mitigations are captured

6) Problem Management & Continuous Improvement

• Detect trends from incident patterns (e.g., repeated throttling, recurring auth failures, dependency outages).
• Raise and contribute to problem records with evidence and recommended preventive actions.
• Participate in post-incident reviews and implement corrective/preventive tasks within the support scope.

Skills required (must-have)

• Working knowledge of AI platforms: OpenAI / Azure AI, common operational failure patterns (auth, throttling, safety filters, latency, deployment errors).
• Strong cloud fundamentals: Azure (preferred) plus familiarity with AWS and GCP (networking basics, IAM, logging/monitoring, secrets management, API services).
• Strong problem-solving and triage skills; ability to isolate whether issue is app defect, platform issue, integration issue, or user/process issue.
• ServiceNow (or equivalent ITSM) proficiency: incident/request/problem workflows, categorization, SLAs, knowledge base.
• Excellent documentation skills: clear, structured runbooks and KB articles.
• Strong written and verbal communication; comfortable collaborating with L1, app owners, platform teams, and L3 engineering/product support.

Experience (typical)

• 4+ years in production support / operations (L2/L3-facing), ideally in cloud-native environments.
• Exposure to AI/ML or AI-enabled applications in production is strongly preferred.

Desirable / nice-to-have

• Familiarity with observability tooling and queries (e.g., KQL/App Insights, Azure Monitor; equivalents in AWS/GCP).
• Basic scripting/automation (PowerShell/Python) to accelerate diagnostics and reduce repeat toil.
• ITIL Foundation or equivalent service management training.

Key measures of success (KPIs)

• Time-to-triage from L1 escalation
• MTTR for L2-resolvable incidents
• SLA compliance (response/resolution) for assigned queue/categories
• First-time-right escalation quality (L3 acceptance rate; reduced ping-pong)
• Knowledge contribution rate (KAs/runbooks created/updated; KB reuse/deflection)
• Recurring incident reduction through problem management inputs

Interfaces & working relationships

• L1 Service Desk: coaching via KB/runbooks; quality feedback on triage and ticket capture
• L3 Product Support (Advisory UK): escalation partner for deep product defects and complex technical issues
• Cloud/Platform teams: support for platform incidents, capacity/limits, identity/networking issues
• App Owners / SMEs: functional clarifications, change/release coordination, operational readiness