AI Consulting

Fixed scope, fixed price, shipped in weeks — by Kevin Kakolla.

I build production AI systems hands-on: agents, voice AI, RAG, LLM evaluation, and fine-tuning. 15+ years of software engineering, 20+ enterprise customers delivered across stacks and domains. Everything below is work I have already done in production — see the build walkthroughs, code, and published models.

Offers

Voice Agent MVP — 2 weeks
Pipecat · Twilio · Deepgram/Cartesia · your LLM of choice
A production-grade phone/voice agent for one workflow (support triage, booking, qualification), with sub-second latency, call recording, and a handoff path to humans. You get the running system, the code, and a loom walkthrough your team can extend.
RAG Pipeline with Evals — 2 weeks
Your docs/tickets/knowledge base · retrieval + reranking · measurable quality
Retrieval-augmented answers over your data with an evaluation harness (golden set, LLM-as-judge, regression tracking) so quality is a number, not a vibe. Deployed in your cloud.
Agent Audit — 1 week
For teams whose agent demos well and fails in production
I instrument your agent, find where and why it fails (grounding, tool-calling, prompt drift, latency), and deliver a prioritized fix list — with the top fix implemented.

Pricing is fixed per engagement and depends on scope — ask. Hourly available for ongoing advisory.

Contact

Email kkakolla@andrew.cmu.edu or DM @nageswarkakolla on X with one paragraph about what you're trying to ship.

© 2026 Kevin Kakolla · Home