
Technical Consultant
ValueLabs, India · Jan 2024 – Present
- Architected a multi-agent incident-response copilot that ingests alerts, metrics, logs, traces, deployment history, runbooks, service ownership records and past postmortems across 12+ microservices to investigate incidents end to end.
- Orchestrated 6 specialized investigation agents through LangGraph, RAG and MCP to correlate evidence, rank root-cause hypotheses, flag probable change triggers and estimate service-level blast radius, cutting analysis time to 15 minutes per incident.
- Delivered human-approved remediation plans, live Slack and Teams incident summaries and source-linked postmortem drafts for 40+ production incidents, integrating Prometheus, Loki, Tempo, Docker, Kubernetes and Azure into a single observability stack.
- Engineered an AI POC-discovery copilot that researches a prospect's business model, product line, technology footprint and public priorities from a single company name or URL, screening 100+ target accounts to date.
- Grounded recommendations in internal delivery artifacts spanning 15+ case studies and reusable accelerators to propose relevant POCs, account stakeholders and next actions for sales and delivery teams.
- Integrated Slack and Teams for conversational account briefs, follow-ups, task creation, meeting summaries and manager-ready progress updates, reaching 20+ account stakeholders on a weekly cadence.
- Developed an AI-powered ATS that parses resumes, matches candidates against role-specific requirements and conducts structured 30-minute L1 voice interviews, processing 500+ resumes to date.
- Designed adaptive interview flows across 25+ competency areas that probe resume claims, project depth and role-specific skills, generating rubric-based evaluations, evidence summaries and follow-up questions for recruiter review.
- Established proctoring review signals and AI-guided L2 interviewer matching based on skills, delivery history, workload and prior feedback, equipping interviewers with L1 context and tailored evaluation prompts for 200+ candidates.
- Built an AI proposal-generation workflow that converts a client brief, RFP, or content idea into an editable, brand-aligned PowerPoint deck.
- Indexed 1,000+ historical presentations to retrieve references, reusable slides and visual patterns across 30+ proposal categories, each with source attribution.
- Generated proposal narratives and slide-wise content with reviewer built into the workflow before final export, reducing first-draft turnaround from roughly two days to under four hours.
- Developed a grounded legal-review assistant for a legal-services client that reads contracts in natural language and extracts obligations, deviations and key commercial terms across 15+ clause categories.
- Identified risky clauses, missing protections and inconsistent provisions by comparing 50+ agreements against approved templates, historical decisions and negotiation playbooks.


