Where serious AI work actually gets built.
A curated marketplace connecting companies with verified AI engineers — through technical proposals, proof of work, and milestone-based escrow.
Companies publish a technical problem, the existing stack, and measurable success criteria — not a vague job post.
Engineers are matched on skill compatibility and proof of work, and every score can be traced back to the evidence behind it.
Funds are held and released only when agreed deliverables are approved. Trust is part of the flow, not a badge.
Briefs currently taking proposals
Each brief states the problem, the environment it lives in, and the numbers that define done.
Adversarial evaluation of a customer-facing LLM assistant
Stand up an adversarial test programme covering prompt injection, data exfiltration and tool abuse, run as a release gate.
- Budget
- ₹5L–₹7L
- Timeline
- 6 wks
- Engagement
- Contract
- Proposals
- 4
- Team
- 1 eng
- Mode
- Remote
- Escrow
- Not funded
- Level
- Staff
Streaming ASR for code-switched Hindi–English contact centre audio
Build a streaming transcription pipeline that holds accuracy on code-switched, noisy telephony audio at 1,000+ concurrent streams.
- Budget
- ₹7L–₹9.5L
- Timeline
- 10 wks
- Engagement
- Contract
- Proposals
- 6
- Team
- 1 eng
- Mode
- Remote
- Escrow
- Not funded
- Level
- Senior
GPU inference cost reduction for sequence model serving
Reduce cost per inference run on a mixed research and production GPU fleet without regressing accuracy beyond an agreed tolerance.
- Budget
- ₹8L–₹11L
- Timeline
- 9 wks
- Engagement
- Contract
- Proposals
- 9
- Team
- 1 eng
- Mode
- Hybrid
- Escrow
- Not funded
- Level
- Principal
Containment-first support agent for 38k daily conversations
Rebuild a tool-calling support agent with explicit escalation boundaries, replayable traces, and measured auto-resolution.
- Budget
- ₹9L–₹12L
- Timeline
- 8 wks
- Engagement
- Contract
- Proposals
- 17
- Team
- 1 eng
- Mode
- Remote
- Escrow
- ₹3L
- Level
- Senior
Engineers with verifiable work
Discovery here leads with shipped systems and measured outcomes, not star ratings.