Graduated from Baylor University with a BS in Biology. Flatiron School SE graduate looking to hopefully advance the field of medicine through programming.
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
portfolio
June 22, 2026 – Present
Personal portfolio site (static HTML/CSS/JS) for Roshaan Singh — AI/ML & full-stack engineer. Deployed on Vercel.
View Projectagent-orchestration
June 21, 2026 – Present
Production-shaped agent runtime: plan/act/observe/reflect + token-budgeted & durable memory + human-in-the-loop + crash-resumable runs. Live demo on HF Spaces.
View Projectrag-citations
June 21, 2026 – Present
RAG with hybrid (dense+BM25, RRF-fused) retrieval and verified citations — hallucinated citations are stripped before output. Live demo on HF Spaces.
View Projectsemantic-cache
June 21, 2026 – Present
Semantic cache for LLM calls: returns stored responses for semantically-equivalent prompts. Exact + NN lookup, TTL/LRU, near-miss calibration. Live demo on HF Spaces.
View Projectmultiagent-critique
June 21, 2026 – Present
Multi-agent evaluator-optimizer loop: a generator + parallel critics iterate until quality passes a rubric. Routing + HITL. Live demo on HF Spaces.
View Projectcost-router
June 21, 2026 – Present
Routes each LLM request to the cheapest capable model (40-60% spend cut). Heuristic classifier, verify+escalate, OpenAI-compatible proxy. Live demo on HF Spaces.
View Projecttxt2sql-guardrails
June 21, 2026 – Present
Natural-language to SQL with an AST-level guardrail that statically blocks destructive queries. 100% block rate on an adversarial corpus. Live demo on HF Spaces.
View Projectdsa-solver
June 10, 2026 – Present
Spec-driven DSA solution engine: brute-force + optimal solutions with complexity proofs, edge cases, and interview coaching
View Projectai-ml-portfolio
June 8, 2026 – Present
AI & ML engineering portfolio — multiple LLM evaluation and safety projects showcasing my AI/ML knowledge and agentic coding abilities
View ProjectCultural Fit Analysis
The candidate's project portfolio is heavily skewed towards AI/ML and Python backend development, which does not align well with a 'Frontend Developer' target role. While there are two frontend-related projects ('portfolio' and 'god-squad-nba'), the depth and quantity are significantly less compared to the AI/ML work. This indicates a potential mismatch in primary interest and expertise for a dedicated frontend role.
Soft Skills & Operational Fit
Insufficient data to assess soft skills and operational fit. Psychometric test scores are 0, providing no insight.