|
AI Engineer & Senior Python Developer with 4+ years of experience spanning enterprise software development at Cognizant, AI model evaluation at Labelbox (Alignerr), CTO leadership at Technitude Info Solutions (managed 15+ engineers, delivered 50+ client projects), and entrepreneurial brand leadership as Founder of Trilimedia Digital Agency. What I Engineer
Open To |
|
| Domain | Proficiency | Details |
|---|---|---|
| LLM Evaluation | Advanced | Model output scoring, safety benchmarking, edge-case identification, evaluation scenario design at $80-100/hr |
| Prompt Engineering | Advanced | Jailbreak testing, system prompt optimization, multi-turn attack vectors, structured output design |
| RAG & Retrieval | Intermediate | Vector databases, document chunking strategies, hybrid search, embedding optimization |
| AI Model Serving | Intermediate | Model deployment pipelines, API wrappers, latency optimization, A/B testing frameworks |
| Data Engineering | Intermediate | ETL pipelines, data processing automation, feature stores, n8n workflow orchestration |
| Enterprise Full Stack | Advanced | Spring Boot + Angular enterprise apps, cloud-native microservices, CI/CD, 80%+ test coverage |
Alexander — Autonomous AI Agent Framework
A production-grade autonomous AI agent framework with three-tier memory architecture (working, episodic, semantic), real-time voice pipeline via VAD+Whisper, and a sophisticated goals scheduler with priority queues. Built entirely from scratch to bridge AI research and deployable engineering.
| Attribute | Detail |
|---|---|
| Stack | Python, FastAPI, WebSocket, Redis, PostgreSQL, Docker |
| Scale | Multi-agent orchestration with async event-driven communication |
| Performance | Sub-100ms inference pipeline; VAD-triggered wake word detection |
| Security | Sandboxed execution, input validation, rate-limited endpoints |
| Impact | Full-featured AI assistant framework ready for production deployment |
| Repository | github.com/ShyamKumar1/alexander |
Three-tier memory architecture, goals scheduler with priority queues, real-time voice interface via VAD+Whisper pipeline, and modular plugin architecture.
AI Code Evaluator — Automated Code Quality Scoring
An automated framework that scores AI-generated code across four weighted dimensions — correctness (40%), edge cases (25%), performance (20%), and code quality (15%) — producing a 0-100 score with detailed Markdown reports. Inspired by Labelbox and Scale AI evaluation pipelines.
| Attribute | Detail |
|---|---|
| Stack | Python 3.11+, pytest, PyYAML, GitHub Actions CI |
| Scale | Multi-dimensional scoring matrix with configurable weight profiles |
| Performance | Batch evaluation pipeline with parallel test execution |
| Security | Sandboxed code execution, input validation, CI-scoped permissions |
| Impact | Production evaluation tool used for AI model benchmarking workflows |
| Repository | github.com/ShyamKumar1/ai-code-evaluator |
100% pytest coverage with CI matrix across Python 3.11-3.13. Configurable rubric weights, detailed scoring breakdowns, and pass/fail thresholds for automated quality gates.
DSS Accounting System — GST-Compliant Silk Industry Platform
A full-stack monorepo accounting system purpose-built for the silk saree industry. Features double-entry accounting, GST engine (CGST/SGST/IGST), invoice management, credit notes, bank reconciliation, and GSTR-1/3B reporting for 5 firms. Turbo monorepo with Prisma ORM and PostgreSQL.
| Attribute | Detail |
|---|---|
| Stack | TypeScript, Turbo, Next.js, Fastify, Drizzle ORM, Turso, Prisma, PostgreSQL, Redis, Docker |
| Scale | Multi-tenant architecture serving 5 firms with firm-level data isolation |
| Performance | Redis-cached queries; sub-100ms report generation for GSTR filings |
| Security | Row-level security, session-based auth, audit logging for all financial transactions |
| Impact | Production system managing real financial data for multiple businesses |
| Repository | github.com/ShyamKumar1/dss |
End-to-end GST compliance engine: auto-calculates CGST/SGST/IGST based on transaction type. Generates GSTR-1 and GSTR-3B returns. Multi-firm dashboard with role-based access control.
Money Machine — Revenue Automation & Lead Engine
A revenue automation platform with Express dashboard, n8n workflow engine, and three active automation pipelines — Lead Gen ($197-497/mo), Content Engine ($147-247/mo), and Booking System ($97-197/mo). Processes 1,000+ qualified leads with real-time dashboards.
| Attribute | Detail |
|---|---|
| Stack | Python, Express.js, n8n, SQLite, REST APIs |
| Scale | 1,000+ leads processed with automated enrichment and scoring |
| Performance | Real-time lead ingestion with sub-minute processing latency |
| Security | API-key authenticated webhooks, rate-limited endpoints, encrypted storage |
| Impact | Revenue-generating system with active client subscriptions |
| Repository | github.com/ShyamKumar1/money |
Three-tier automation pipeline: lead sourcing & enrichment → content generation & distribution → booking & follow-up automation. n8n orchestrates the entire workflow with Hermes handling data processing.
Trilimedia Brand Projects — Tropica Café & TARA Jewelry
Flagship brand identity and content strategy projects delivered through Trilimedia Digital Agency. Tropica Café was repositioned as a premium café brand with full visual identity and 6-month content system. TARA Jewelry received a complete brand identity suite with AI-augmented documentation and structured launch architecture.
| Brand | Deliverables |
|---|---|
| Tropica Café | Brand repositioning, visual identity system, 6-month content strategy, social media framework, story-led engagement model |
| TARA Jewelry | Full brand identity suite, logo system, design templates, storytelling assets, structured launch campaign architecture |
Both projects delivered through Trilimedia Digital Agency (founded 2023). Combined result: measurable customer engagement lift and zero off-brand incidents across all accounts.
Hermes Vault — Automated Knowledge Management
An automated knowledge management and archiving platform that captures, organizes, and surfaces daily engineering insights, project artifacts, and research findings through an elegant web interface. Runs on a nightly cron-driven build pipeline.
| Attribute | Detail |
|---|---|
| Stack | TypeScript, Next.js, Tailwind CSS, Markdown, Netlify, GitHub Actions |
| Scale | Automated daily ingestion with cron-driven builds |
| Performance | Static site generation; sub-second page loads globally via CDN |
| Security | API-gated write operations, HTTPS everywhere |
| Impact | Centralized knowledge base serving as daily engineering journal |
| Repository | github.com/ShyamKumar1/hermes-vault |
End-to-end automated pipeline: GitHub Actions cron triggers content ingestion → Markdown generation → static site build → Netlify deployment. Sections for daily logs, blogs, projects, contracts, and archive.
Senior Python Developer — AI Model Review-8B5CF6?style=flat-square&logo=windowsterminal&logoColor=white){kind=icon}
Jun 2026 — Present · Remote
|
Chief Technology Officer (CTO)
Aug 2025 — Jan 2026 · India
|
Founder & CEO
2023 — Present · India
|
Programming Analyst
Mar 2022 — Nov 2024 · India
|
| Recognition | Details |
|---|---|
 |
Cognizant award for exceptional contribution, delivery quality, and team collaboration |
 |
Delivered as CTO at Technitude — maintained 99.9% uptime across all deployments |
 |
Contributions to Marimo, LM Studio Bridge, OpenCode, and Hermes Agent ecosystem |
 |
Active Labelbox/Alignerr contractor evaluating frontier LLMs at $80-100/hr |
 |
Founded Trilimedia Digital Agency — 10+ clients, zero off-brand incidents |
-8B5CF6?style=for-the-badge&logo=amazonwebservices&logoColor=white){kind=icon}
learning:
- Advanced LLM evaluation methodologies
- Distributed systems design patterns
- AI safety and alignment research
building:
- Autonomous AI agent frameworks (Alexander)
- Revenue automation pipelines (n8n + Hermes)
- Production-grade AI evaluation tooling
exploring:
- AI red teaming and bug bounty programs
- Multi-modal AI architectures
- Brand strategy + AI-augmented content systems
open_to:
- Senior AI Engineer / Python Developer roles
- Technical consulting and fractional CTO engagements
- Open source collaborations
- Speaking and community contributions
"Engineering is not just about writing code — it's about architecting systems that make a difference."