Independent ventures

Built end to end. Then red-teamed.

Four products I designed, architected, coded, and operated solo, from concept to deployment. Each owns a hard problem: knowledge-graph modeling, multi-provider LLM orchestration, graph algorithms, or the discipline of autonomous agents. Together they are proof that my leadership comes from doing the work, not describing it.

Four mechanisms
01

Ondoway

A local storyteller in your pocket.

A Neo4j knowledge graph of city points-of-interest and narrative beats that powers GPS-triggered, personalized audio walking tours. The launch city is Paris. The pitch is simple: a knowledgeable local, narrating the right story at the right corner.

What it proves  Senior-architect graph-modeling judgment, including a documented record of what the design deliberately skips (RDF/SPARQL, OWL, embeddings) and the triggers to revisit each. I owned the graph, the backend, the ML and audio pipeline, the generative content, an interactive authoring frontend, and a Flutter mobile app with GPS geofencing.

Graph model

Versioned, event-centric, CIDOC-CRM-aligned schema. Eight core entity types and twelve relationship types, several as reified edges carrying their own properties. Controlled vocabularies, canonical IDs, per-fact provenance. A lens taxonomy of eight parent genres and twenty-one sub-genres.

Stack & AI

Python 3.11 and FastAPI (JWT, Google and Apple sign-in, rate limiting), Neo4j 5.28 with spatial indexes, Flutter mobile, AWS S3. Dual-provider TTS behind a protocol, narration glue via Claude Haiku, fact-check via Gemini.

Scope & rigor

Sixteen REST endpoints, a 25-module deterministic tour engine in two modes, a browser graph editor, and a Leaflet editorial workbench. 177 tests with a CI tour-grading regression gate. Sentence-level provenance: every spoken sentence traces to a beat or a whitelisted label, so AI cannot fabricate.

02

Persuaider

Practice the hard conversation.

Admins author negotiation scenarios with in-character AI personas. Learners practice live. An LLM evaluates each session and returns scored, framework-based feedback, with built-in CLEAR and AIDA frameworks plus admin-defined weighted ones, a 0-to-100 score, and the winning arguments.

What it proves  Production-grade multi-provider LLM orchestration, not a single-API wrapper. Genuine full-stack ownership, and real operational maturity, including a “zero skipped tests” rule and a dedicated security and red-team layer.

LLM orchestration

A three-tier chain with pinned versions: Claude Sonnet 4.5 to Gemini 2.0 Flash to GPT-5.2 Instant. Exponential backoff, retryable-versus-terminal error classification, transparent mid-conversation failover. The LLM plays three roles: persona simulation, rubric evaluation, and scenario generation.

Stack

TypeScript and Next.js 16 (App Router), React 19.2, Tailwind v4. Prisma 6 over SQLite in dev and PostgreSQL in prod, eleven data models. Clerk auth, Upstash Redis rate limiting, React-PDF export, hosted on Vercel at roughly $0.07 per session.

Scope & rigor

Roughly 12,875 lines of non-test TypeScript, 24 React components, 65 API routes, fourteen documented user pathways across three actor types. 1,246 Jest unit tests and 15 end-to-end, zero skipped, including 67 security and middleware tests and dedicated red-team files.

03

Lore / DependIQ

The verification layer AI coding agents lack.

A multi-tenant web app that builds a cross-project dependency graph over a workspace of repos and computes the blast radius of a change. Given a package upgrade, it shows which projects across the org break, in dependency order, with a streaming LLM explanation of why.

What it proves  Real graph-systems and graph-algorithm work, deliberate polyglot persistence with explicit reasoning about the dual-write atomicity gap, and a strong leadership signal. The codebase ships alongside a brutally honest senior-engineer audit of itself, with file-and-line evidence and a prioritized fix plan, plus an investor-grade competitive memo.

Graph algorithm

Hand-written Cypher that starts from a package, finds directly-affected projects, then traverses inter-project semantics up to five hops via shortest path. Results carry distance and impact type, severity-bucketed by hop distance, behind a one-hour cache. Design-target scale near 50,000 package nodes.

Stack

Python 3.11 and FastAPI. Polyglot persistence: PostgreSQL 14+ for identity, workspaces, and history, plus Neo4j 5.x with APOC for the dependency graph. WorkOS AuthKit. A deliberately build-free frontend in HTMX, Jinja2, Alpine, and Tailwind. Hosted on Render.

LLM routing

Provider-agnostic via litellm with a custom model router. Eleven task types tiered across Claude, OpenAI, and local Ollama models, in four modes: Quality, Balanced, Cost, and Local-Only for privacy, each with its own failover chain. Over 200 tests.

04

Crucible (claude-kit)

Discipline for AI coding agents.

An open-source kit that encodes disciplined engineering practice into an AI coding agent: slash-command skills, multi-perspective agent definitions, reusable patterns, and mechanically-enforced behavioral hooks. Ten skills, ten agents, four patterns.

What it proves  A clear thesis, in working code. An LLM coding agent must be treated as an adversary and disciplined with code, not prose, applying adversarial verification, test-first practice, deterministic gating, and honesty contracts. It is the operational twin of the management research.

The forge skill

A multi-perspective answer engine with hostile, evidence-gated verification behind a deterministic gate. Parallel generators, per-claim skeptics that re-open each cited file-and-line and confirm the quoted bytes, up to three adversaries. A pass requires positive evidence, never the mere absence of objections.

What ships

Skills as Markdown with YAML frontmatter, a Node verification engine, and a Bash installer with enforcement hooks. The flagship workflow is backed by thirteen committed Node test files. Ten agents each carry an authority rank and persistent memory, and conflicts merge in authority order.

More skills

build-fix, an iterative build-fixer that never skips a failing test to go green. Plus ship, retro, error-diagnosis, prwalk, and agent-management skills, and four runnable patterns: observe, mutate, verify, and pipelines.

Earlier and notable

A longer trail of systems, from Carnegie Mellon research to weekend builds, that show the same instincts at smaller scale.

Amazon ML Hackathon OpenCV image-matching to detect whether two warehouse bins held the same objects. An ensemble of feature-matching and edge methods with averaged confidence, for a real Fulfillment and Kiva Robots problem.
N-an-Instant A custom Google Instant built from n-grams over a 10,000-book corpus and Bayesian statistics, with Hadoop into HBase and a jQuery autocomplete front end.
End-to-end AWS cloud service The second most cost-effective RESTful API in the cohort, analyzing a 200 GB Twitter corpus with an Elastic MapReduce ETL into HBase.
Twitter hashtag similarities Co-occurrence counts over a 50 GB corpus via Hadoop, with a 78% speedup over the default solution.
Wikipedia page-views timeline Monthly view counts via Elastic MapReduce on a twenty-dollar operating budget.
Embedded & systems work An energy-aware real-time embedded Linux kernel, a Raspberry Pi motion-sensing capstone, and a discrete-time elevator simulation over a CAN network.
For investors & design partners

Talk about the ventures.

Open to conversations about these products — the architecture, the code, or where each one goes next.

Start a conversation GitHub The research behind them
Sairam Krishnan
Engineering leadership · AI & ML
Apple · Wharton MBA 2026
Pages
The record
Reach me