Future releases of PolySwarm will be based on changing threat market dynamics and actual web usage. Based on our observations of PolySwarm usage, we see the following features as opportunities for the network: flexible "artifact" confidentiality; Basic: "artifacts" are selectively exposed to specific experts and SaaS service providers; Further: "artifacts" are never exposed to anyone outside the SGX enclaves of the fully homomorphic engine; New tools (e.g., subscribing to ready-made threat news feeds); Provides support for working engine containers for endpoints, replacing the current endpoint protection suite.

The second stable version of PolySwarm has two main goals: (1) expand "bids" and "rewards" to handle a wider variety of "artifacts" (e.g., URLs and Web streams) and (2) ease business and personal usage of PolySwarm. The new PolySwarm version should be able to handle many more "artifacts" such as automatic detection of spooks by placing "rewards" and "offers" directly on URLs and web content. Extending end-user protection will be a priority for PolySwarm once stable development is complete. We're looking forward to developing an open source reference example of an endpoint protection suite that automatically blocks malicious "artifacts" with PolySwarm intelligence, and we think this open proxy framework is key to widespread adoption.

Prior to moving from the Warm test network to the true PolySwarm market, the focus was on cleaning up bugs in PolySwarm Gamma. If time/money allows, the larger goal of PolySwarm1.0 Stable is to introduce high level "artifact" secrecy and only release "rewards" to limited participants.

The focus of The PolySwarm Gamma release will be on delivering mediating capabilities that facilitate collaboration between representatives, experts, and mediators, who will be able to encapsulate their expertise into a "work engine" for specific types of artifacts. We expect them to advertise these "working engines" in the PolySwarm registry, which will allow representatives to use the "working engines" in a risk-free way until the release of PolySwarm1.0 (version Gamma will continue to use the PolySwarm testnet). In addition, version Gamma will consider enhancing and extending the reward tool based on feedback and experience from Alpha and Beta. For example, we would expect this to be beneficial to the "reward" system: the ability to solicit claims that are kept secret from other "experts" (confidentiality) and only revealed after the "claim" expiration date. We expect the PolySwarm Gamma release to accomplish the following goals: Worker Description Language (WDL) that describes artifact analysis capabilities and attribution of expert copyright (for reputation); Distributed, searchable work engine registration system and related functions; Confidentiality of "reward claims" prior to the expiration date; Complete the selection of "Arbiter", response, and end-to-end "reward" feedback loop; We expect release Gamma to have full market functionality.

PolySwarm Beta will focus on delivering "PolySwarm Bid" support through the Thunderbolt Network Channel. Bids are designed to facilitate a higher volume of artifact evaluation than rewards, which also requires more efficient settlement of each transaction on the blockchain. The Smart-contract version of PolySwarm Beta is expected to be implemented as a direct, peer-to-peer thunderbolt network channel between delegates and experts (using a test network). We expect The PolySwarm Beta to accomplish the following: Bid channel; Establish distributed "channels"; "Channel closed" settlement NCT; Reference to generate and respond to "bid" execution instances; Collect "fees" and dispatch smart contracts; Research and quantify the trust required for a specific amount of "artifacts" (as NCT); Offer/Reward development Kit: "experts" and "representatives" automatically propose "reward" and "offer" test kits and implementation frameworks.

Following token sales, the focus will be on delivering a prototype of the "Reward" end-to-end workflow, this initial build will kick-start the Active development process of the PolySwarm team, where "representatives" and "experts" will be able to trade through the PolySwarm test network (" Testnet "). A basic Reward smart contract will be available on the PolySwarm beta network, allowing delegates to post rewards, experts to make claims, and Delegates to make judgments. We expect PolySwarm Alpha to achieve the following goals: Reward Manager Smart contract to save reward amounts and fees, accept claims, deliver verdicts, reward the right "experts"; Non-polyswarm storage for "artifact" publishing and extraction (e.g. S3 and/or IPFS); The prototype algorithm used to select the "arbitrator" when the "reward" expires; Refer to the "reward" announcement, the "assertion" response, and the "arbiter" to determine true facts for example.