Shouryam Kumar
AI Consultant and Product Strategist who identifies high-impact business problems, designs GenAI solutions, and delivers measurable operational outcomes. Proven track record across moderation automation, LLM evaluation, AI safety, and personalisation systems — consistently translating technical AI capabilities into decisions stakeholders can act on.
Targeting AI Analyst, AI Transformation Consultant, and Technical Business Analyst roles.
Experience
Co-founder & CTO
Ambrix Lifemedia
Remote
- Analysed content moderation as a critical scaling bottleneck; designed and operationalised an LLM-powered automation solution with safety guardrails and fallback handling, reducing manual review effort by 80% and eliminating the need for additional moderation headcount.
- Identified gaps in AI output reliability; led structured development of a quality assurance framework covering evaluation criteria, test-case design, and failure-mode analysis — standardising AI governance across the product delivery lifecycle.
- Translated business requirements into AI solution designs, balancing operational constraints, user experience, and implementation complexity across the product roadmap.
AI Founding Engineer
DailyFitness.ai
Remote
- Gathered and analysed personalisation requirements; designed an LLM-powered recommendation solution using dynamic context routing, accelerating user engagement through relevant, context-aware content delivery.
- Diagnosed latency as the primary barrier to product responsiveness; evaluated infrastructure options and redesigned caching architecture, reducing API response time by 40% and cutting infrastructure costs.
- Assessed LLM output quality risks across safety and accuracy dimensions; designed evaluation standards and human-in-the-loop validation checkpoints that measurably reduced hallucinations and unsafe outputs in production.
Technical Community Manager (DevRel)
Okto
Remote
- Conducted structured discovery of developer onboarding friction, identifying and prioritising 15+ pain points; partnered cross-functionally with product and engineering to implement targeted fixes that measurably improved developer activation rates.
- Evaluated manual support workflows for inefficiency; automated ticket routing and team notifications using Python and Slack API, cutting issue resolution time by 50% and redirecting team capacity to higher-value work.
Operations Lead
The Phoenix Guild
Remote
- Designed and delivered structured technical programmes for 1,500+ developers across 10+ regions, achieving 90%+ satisfaction by standardising content delivery and adapting execution to diverse stakeholder needs.
- Defined programme governance standards and delivery frameworks across distributed teams, ensuring consistent execution and stakeholder alignment across cross-cultural, multi-region operations.
Education
Lakshmi Narain College of Technology
Bachelor of Technology
Skills
Requirement analysisstakeholder managementsolution designprocess optimisationworkflow automationAgile deliveryproduct roadmapstructured problem-solvingLLM evaluationAI risk assessmentAI safetyhuman-in-the-loop evaluationprompt optimisationresponsible AINIST AI RMFstructured testingRAG pipelinesagent orchestrationtool callingprompt engineeringLangChainLangGraphOpenAI SDKLangSmithHuggingFaceExperiment designA/B testingregression analysispandasscikit-learn
Projects
LLM Evaluation & AI Governance Framework
- Designed a structured AI evaluation system assessing LLM outputs across accuracy, hallucination rate, toxicity, and safety; operationalised structured test datasets, automated scoring pipelines, and adversarial red teaming protocols to support production AI risk assessment.
- Produced prioritised evaluation reports mapping model failure modes to business risk, enabling data-driven deployment decisions and reducing post-launch reliability incidents.
Fingle - AI-Enabled Social Platform
- Led end-to-end product delivery of a social platform (feeds, messaging, short-form video) from concept to 400+ active users; designed AI personalisation workflows and resolved critical infrastructure instability, delivering improved platform stability and user retention